ARCH221 URBAN STUDIES, 2024-2025 Essay Topics The essay accounts for 70% of your total module mark. Please choose one topic from the list below. Note that the bullet points under each topic are intended as prompts to help you develop your ideas, not as a suggested structure for the essay. You should not simply answer the questions in the bullet points; instead, use them to guide your thinking. Essay length: 2000 words including references (both in text information and reference list/bibliography at the end) (10% flexibility allowed). Online Submission Date: Friday 16th May 2024, 2.00pm via CANVAS. No Hard Copy required. For learning difficult students: Please include a cover sheet stating I confirm that I have a Student Support Information Sheet (SSIS) which recommends ‘Appropriate marking for spelling and written expression for students with Specific Learning Difficulties’. Topic 1: Explore the historical forces that have shaped an urban area and discuss how design can respond to this historical development in a positive manner · Investigate the natural, social, economic, and political factors that influenced the development of a specific urban area (e.g., a neighbourhood, district, or city) of your choice. · Discuss how these factors shaped the urban area and its main characteristics as a result; · Explore how urban design can respond to these characteristics and enhance the area’s identity; Draw lessons can be learnt. Topic 2: To what extend should a city be planned or not planned by architects/planners? · Choose a planned city or a planned district within a city by a particular architect/planner or a team of architects and planners as your case study; · Considering the arguments related to the top-down vs bottom-up; grand planning products vs piecemeal development. · Discuss how the planned city works for its residents in the long run; · Discuss the role of architects and planners in the design of cities and lessons can be learnt. Topic 3: Can an urban design principle be applied to a context different from where it originated? · Choose a specific urban design principle discussed in the lecture (e.g. mixed use, legibility, diversity, ease of movement, robustness) as the focus of your investigation; · Select a city or town you are familiar with as a case study. Investigate whether this urban design principle is relevant to the chosen city. If not, explain why; If yes, discuss how the principle can be applied to the local context. · Give your critical view on how the principle could better guide design and planning in the case study city or town. Topic 4: How to regenerate Liverpool’s Chinatown through urban design? · Explore Chinatown’s history and how urban design can celebrate its cultural heritage, focusing on landmarks, buildings, businesses, and festivals. · Discuss its public spaces and how to involve the local community in the design process. · Examine connectivity and accessibility of Chinatown, enhancing pedestrian, cycling, and public transport links. · Consider green design strategies to boost environmental sustainability of Chinatown. · Focus on one or two aspects (or a topic of your choice) and avoid covering everything. You may link your essay to your ARCH252 studio design. Topic 5: How does technology transform. urban spaces and exert changes on humankind? Amin and Nigel have explored this question in their book Seeing like a City (2016). However, as geographers, they adopted a more macroscopic view. How can we, as architects and urban designers, unpack this tension at a human scale and in a more visually engaging way? · You are invited to use a film or fiction to illustrate the exaggerated impact of urbanisation, technology and digitalisation on humankind. · Suggested films included, but not limited to: Matrix (1999-2003, 2021), Blade Runner (1982, 2017), Ghost in a Shell (1995, 2017), Aeon Flux (2005), Equilibrium (2002), Brazil (1985), Stalker (1979), or classics like Metropolis (1927). · Choose one or two scenes depicting urban spaces in these films as examples. Analyse these spaces and actions using the suggested text and other literature. Avoid attempting to cover the entire city featured in the films. The Prologue and Chapter 3 in Seeing like a City will be the particularly relevant. · Consider how the observations allow you to re-evaluate existing cities and urban life. Be speculative but also evidence-based. For example, could Blader Runner (set in a futuristic Los Angeles with its urban grid) be relevant for understanding Beijing, its urban structure, and its vast population today and in the future? You may refer to other lectures to address this part. Topic 6: What can we learn from the evolution of housing estate landscapes over time? · Select a post-war housing estate landscape in a city you know well and ideally can visit; · Research the history and changes to the post-war housing estate landscape. What were the main design elements and consideration? · Compare the historical ideas to today’s situation: What remains? What has been lost? How has the estate changed over time? · Compare and discuss the shifts between the original intentions and the current situation. Who were the intended audiences, and who lives there now? · Discuss critically what we can learn from the original design and subsequent evolution of your chosen landscape. Topic 7: How can we create inclusive public spaces for underrepresented social groups in our cities? · Focus on a specific social group in this essay, such as children, the elderly, women, ethic minorities or others. · Observe their use of public spaces on a daily basis or during events, and discuss how the physical and management features of the space affect their access and experience. · Choose one or two case studies that you can visit and analyse in detail. · Address the topic question and draw lessons from your findings. Topic 8: How can urban planning help mitigate the adverse effects of climate change in your city? · Focus on a specific aspect of the adverse effects of climate change, such as rising temperatures, urban heat island effect, flooding and sea level rise or others. · Choose a particular city or district to discuss how planning aims to mitigate these effects. Examine the strategies proposed, the extent of their implementation, and the reasons behind their success or limitations. · Evaluate how successful these planning strategies are and explain why. · What lessons can we learn from this case study? Topic 9: What impact does urban design have on energy consumption in the built environment, regarding climate change? · Consider the factors that contribute to urban heat islands. · Reflect on the combined future challenges posed by climate change and urban heat island effect. · Examine different urban layouts in various climatic regions and explore why they are designed as they are. · Use one or two case studies, inspired by the lecture but extending beyond the lecture, to demonstrate the effects of greening in city centres. · What lessons can we learn from the case study? Naming convention of your submitted file The essays will be marked anonymously by a group of assessors so please make sure you don’t present your personal information anywhere in the submitted file (you need to include a cover sheet if you have an appropriate marking arrangement in place). Please name your file in the following format ‘Topic x-your own short title’. In the file name, you must indicate which topic your essay addresses and give a short title for it. You need to upload the file in Word or PDF formats to CANVAS. Writing style. and referencing The essay should be written in an academic style, using clear language. It should have a clear structure that includes an introduction, background and literature review, analysis or case studies, discussion and conclusion (refer to the Marking Criteria document on CANVAS). All arguments should be well justified, and illustrations used as appropriate. There is no specific requirement for font, text size and figures. The format of the essay should be consistent and legible. Sources or citations from literature should be properly referenced in a consistent manner. Please consult the Code of Practice on assessment Appendix L Academic Integrity Policy for definitions of dishonest academic practice, plagiarism and relevant penalties https://www.liverpool.ac.uk/media/livacuk/tqsd/code-of-practice-on-assessment/appendix_L_cop_assess.pdf . We suggest The Chicago Style. Author-Date system to be used as your reference style. The following examples illustrate citations using the author-date system. Each example of a reference list entry (the first line) is accompanied by an example of a corresponding parenthetical citation in the text (the following line). These examples are citations from book(s), book chapters, journal articles, and websites. For other types of citations, please refer to http://www.chicagomanualofstyle.org/tools_citationguide.html . Book(s) Ward, Geoffrey C., and Ken Burns. 2007. The War: An Intimate History, 1941–1945. New York: Knopf. (Ward and Burns 2007, 52) Books or articles with four or more authors List all of the authors in the reference list; in the text, list only the first author, followed by et al. (“and others”): (Barnes et al. 2010) Chapter or other part of a book Kelly, John D. 2010. “Seeing Red: Mao Fetishism, Pax Americana, and the Moral Economy of War.” In Anthropology and Global Counterinsurgency, edited by John D. Kelly, Beatrice Jauregui, Sean T. Mitchell, and Jeremy Walton, 67–83. Chicago: University of Chicago Press. (Kelly 2010, 77) Article in a print journal In the text, list the specific page numbers consulted, if any. In the reference list entry, list the page range for the whole article. Weinstein, Joshua I. 2009. “The Market in Plato’s Republic.” Classical Philology 104:439–58. (Weinstein 2009, 440) A citation to website Content can often be limited to a mention in the text (“As of July 19, 2008, the McDonald’s Corporation listed on its website . . .”). If a more formal citation is desired, it may be styled as in the examples below. Because such content is subject to change, include an access date or, if available, a date that the site was last modified. In the absence of a date of publication, use the access date or last-modified date as the basis of the citation. Google. 2009. “Google Privacy Policy.” Last modified March 11. http://www.google.com/intl/en/privacypolicy.html. McDonald’s Corporation. 2008. “McDonald’s Happy Meal Toy Safety Facts.” Accessed July 19. http://www.mcdonalds.com/corp/about/factsheets.html. (Google 2009) (McDonald’s 2008)
COMP3221 Assignment 2: Blockchain Due: Friday, Week 11 by 23:59 1 Introduction In this assignment, you will implement a peer-to-peer (P2P) blockchain system. A blockchain is a chain of blocks (like pages in a ledger), distributed across multiple nodes. Each block contains a set of transactions, which represent changes in ownership of digital assets or other logged information. By completing this project, you will create a network of blockchain nodes that communicate and cooperate to maintain a consistent distributed ledger. 1.1 Learning Objectives • Understand key blockchain concepts and terminology by implementing them in code. • Learn how a P2P network model operates and how to design communication between distributed nodes. • Gain experience in developing a concurrent networked application (with sockets and threads). • Apply principles of crash fault-tolerant consensus in a practical setting. • Improve programming skills in networking and distributed system design, with a focus on blockchain mechanics. 2 Program Structure and Implementation Requirements You may implement your solution in any programming language of your choice. However, your submission must include a wrapper script named Run .sh that serves as the single entry point. For example, if you choose Python, your wrapper script. might look as follows: 1 #!/bin/bash 2 python3 your_solution.py "$@" You are not permitted to use any libraries outside the core libraries provided for your pro- gramming language. For example, if you were to use Python, you may use libraries such as socket, but not networkx. Using any disallowed libraries will result in an immediate 0 for this assignment. If you are unsure if a specific library is allowed, please ask on Ed. You are also permitted to use a build script that compiles your program. This script must be called Build .sh, and will execute before any of the test cases are run on your program. You are permitted to use any libraries that come built into your chosen language, as well as a select range of extra libraries. For the list of allowed libraries, please see Appendix A. 3 Assignment Task You will implement the software for a single blockchain node (peer). Running multiple instances of your node program will form a P2P network. Each node must act both as a server (listening for incoming requests from peers) and as a client (initiating requests to other peers). Nodes communicate by exchanging structured network messages over TCP connections. The ultimate goal is for all non-faulty nodes in the network to reach consensus on a sequence of blocks, thereby maintaining a consistent blockchain. Every node will maintain a TCP server on a specified port to accept peer connections. It will also initiate outgoing connections to all other known peer nodes. The network topology is fully connected: each node should have an independent, long-lived TCP connection to every other node in the peer list. Nodes must handle connectivity issues gracefully – for example, if a connection drops or a peer crashes, the node should detect this (via timeouts) and continue operating with the remaining peers. Each node operates concurrently with multiple threads, for example: • Server Thread(s): Accept incoming TCP connections from peers. For each connected peer, use a dedicated handler (thread) to receive and process messages from that peer. • Consensus (Pipeline) Thread: Continuously monitor the node’s transaction pool and coordinate the block proposal and consensus process (detailed below in Section 8). All nodes in the network will run the same consensus protocol in synchronous rounds to agree on the next block to append. By the end of each consensus round, every correct node should have decided on the same new block to add to its blockchain (assuming not too many nodes have crashed). 3.1 Transaction Pool and Validation Incoming transactions (from clients or other nodes) are collected in a transaction pool (mempool). The node will store transactions in this pool temporarily until they are included in a new block via the consensus process. Each transaction is a JSON object containing spe- cific fields (see Section 5.2). Upon receiving a transaction (via a peer message), the node must validate it before adding it to its pool. Invalid transactions must be explicitly re- fused. The validation rules are: • Correct Nonce Sequence: Each transaction has a nonce field, which must equal the number of transactions from the same sender that have already been confirmed on the blockchain (i.e. the sender’s transaction count in the current chain state). For example, if a sender has 5 transactions in the blockchain (nonce 0 through 4), the next transaction’s nonce must be 5. If a received transaction’s nonce is not exactly one greater than the sender’s confirmed nonce (or 0 if the sender has none yet), then the transaction is out-of-order or duplicate and must be rejected. There must be no two different transactions with the same sender and same nonce in the pool. (The nonce is monotonically increasing per sender, starting from 0.) • Signature Verification: Every transaction is signed by its sender. The signature field must be a valid Ed25519 signature over the transaction’s contents, and it must correspond to the given sender public key. The node must verify the signature using the sender’s public key. If the signature is invalid, the transaction is invalid and should be rejected. • Field Formats: All transaction fields must adhere to the specified format. If any field is malformed (e.g. wrong length or containing non-alphanumeric characters), the trans- action is invalid. When a transaction is rejected due to any of the above rules, the node should not include it in its pool or any future block. Only transactions that pass all validation steps are pooled. 3.2 Block Creation and Addition Periodically, the node will create a new block containing a batch of pending transactions from its pool and attempt to append this block to the blockchain via the consensus protocol. A node proposes a new block when its transaction pool becomes non-empty (triggering an attempt to form. the next block). Each block is a JSON object with the following keys (see Section 5.2 for full schema): • index: The block’s position in the chain (1 for the genesis block, 2 for the first block after genesis, etc.). • transactions: A list of transaction objects, each containing the fields sender, message, nonce and signature. • previous_hash: The current_hash of the previous block in the chain (for the genesis block, a predefined value is used – see Appendix B). • current_hash: The SHA-256 hash of this block’s content (computed over the string representation of the block’s other fields, with keys sorted lexicographically for consis- tency). When a node creates a block proposal, it must set the index to (current blockchain length + 1), include a selection of transactions from its pool, and set previous_hash to the hash of its latest block. It then computes the current_hash for the new block. This block will then enter the consensus protocol to decide whether it will be the next block appended to the chain. After a consensus round (see Section 8), the nodes will have agreed upon one block to com- mit. When a node decides on a block for a given round, it appends that block to its local blockchain. The node should then update its state accordingly: • Execute/confirm the block’s transactions (e.g. increment nonce counters for senders). • Remove the block’s transactions from the transaction pool (since they are now con- firmed on-chain). • If any other pending transactions in the pool became invalid due to this new block, those should be removed as well. For example, if multiple transactions from the same sender were in the pool and one of them was included in the block, any other transaction from that sender with the same nonce or an out-of-order nonce should be dropped. The blockchain should always remain valid: a decided block’s previous_hash must match the hash of the previously decided block. All nodes start with the same genesis block, so if the consensus protocol works correctly, every node’s chain will evolve identically over time (aside from any blocks that crashed nodes failed to append). 4 Network Protocols and Message Formats All network communication between nodes uses a simple message protocol over TCP. Nodes exchange messages that are length-prefixed and encoded in JSON. Each message sent over the network is preceded by a 2-byte integer length field (unsigned, big-endian byte order) that specifies the number of bytes in the message body. The message body immediately follows the length field and consists of a JSON-encoded object. The maximum message length is 65535 bytes (0xFFFF), as the length is 16-bit. This framing allows the receiver to know how many bytes to read for each message from the continuous TCP stream. The message body is a JSON object with two top-level keys: "type" and "payload". The "type" field is a string indicating the kind of message, and "payload" contains the data relevant to that message. There are two types of messages in this protocol: • "transaction" — used to submit a transaction to a node. • "values" — used during consensus to exchange proposed block values (hence “val- ues” refers to proposed blocks). 4.1 Transaction Message Format A message with "type": "transaction" carries a single transaction to be processed. The payload in this case is a JSON object representing the transaction. The transaction JSON has the following keys: • sender (string): The hex-encoded 32-byte Ed25519 public key of the sender (i.e. the account initiating the transaction). This is a 64-character lowercase hexadecimal string. This public key uniquely identifies the user/account. • message (string): Arbitrary UTF-8 text authored by the sender. The consensus protocol treats this as opaque data. The message contains up to 70 alphanumeric characters and spaces. • nonce (integer): A number indicating how many transactions the sender has already had confirmed in the blockchain before this one. The nonce starts at 0 for each sender’s first transaction, and increments by 1 with each confirmed transaction. This field is used to ensure each transaction is unique, and for a given sender (preventing replay attacks or double-spending). • signature (string): The hex-encoded 64-byte Ed25519 signature of the transaction. This signature is generated by the sender’s private key over the contents of the transac- tion (the transaction fields sender, message and nonce concatenated in this order). It is 128 characters in hexadecimal. The signature proves the authenticity of the trans- action (that it was indeed created by the holder of the private key corresponding to sender). When a node receives a "transaction" message, it should parse the payload as a trans- action dictionary and run the validation rules (nonce check, signature verification, etc.) as described in Section 2.1. After processing, the node should return a response to the sender (outside the scope of this message format, a simple True or False as mentioned). The original message does not itself demand a response in JSON format, but the sending function will expect a boolean acknowledgement. Example — Transaction Message: Suppose a user with public key "a578 . . .5363b" wants to log the message "Never Gonna Give You Up" with nonce 0. The user signs the trans- action, producing signature "142e . . . 60c7fd0c". The network message sent to the node would have the following JSON body (length prefix omitted here for clarity of the output): 1 { 2 "type " : "transaction " , 3 "payload " : { 4 "sender " : " a57819938feb51bb3f923496c9dacde3e9f667b214a0fb1653b6bfc0f185363b " , 5 "message " : "Never Gonna Give You Up " , 6 "nonce " : 0 , 7 "signature " : " 142e395895e0bf4e4a3a7c3aabf2 ..... (and so on) " 8 } 9 } The first two bytes of the actual transmission would be the length of this JSON string in bytes (in big-endian). The node, upon receiving these bytes, will reconstruct the JSON and interpret it accordingly. The node’s response to this message (over the same TCP connection) should be a standalone JSON value true (if accepted) or false (if rejected), or simply the literal True/False as a Python boolean serialised. 4.2 Block (Values) Message Format During the consensus protocol, nodes exchange messages of "type": "values". These messages carry block proposals (or requests for them). The "payload" for a "values" message is a JSON structure that can vary depending on context: • In our implementation, every "values" message carries the sender’s own current block proposal (a list containing its block, or an empty list if it has none yet). Send- ing your proposal thus implicitly requests that the peer reply with its own "values" message containing that peer’s proposal. • When responding or broadcasting values, the payload is a list of block objects. Each block object in the list is a JSON dictionary with the fields described above (index, transactions, previous_hash, current_hash). In our consensus algorithm (a single round to decide one block), each node will have exactly one block proposal per round. Thus, you may often send or receive a "values" message where the payload list contains either one block (the peer’s proposal) or multiple blocks (if a node bundles all proposals it knows). In a fully synchronous broadcast approach, a node might send its own proposal to everyone (payload list of one), and also later send a list of all proposals it received (payload of many) — but our simplified approach mainly uses one exchange of single blocks. Every block in the payload list should be formatted as valid JSON with its keys and values as described. For completeness, here’s a breakdown of a block object (these apply whether the block is sent in a values message or printed to stdout): • index: (integer) Block’s position in chain. • transactions: (array) List of transaction objects included in the block. • previous_hash: (string) The hash of the previous block in hex. • current_hash: (string) The hash of this block in hex. Example — Values Request: Node A wants to request Node B’s proposed block for the current round. Node A sends a message to B: 1 { "type " : "values " , "payload " : [ Block_A ] } and receives the following response: 1 { "type " : "values " , "payload " : [ Block_B ] } This indicates a request. When a node receives a values request (empty list), it must respond with a values message whose payload is its own current proposal (or [] if it has none). Example — Values Response (Single Block): Node B has a block proposal (say Block_B) ready. It responds to A with: 1 { 2 "type " : "values " , 3 "payload " : [ 4 { 5 "index " : 2 , 6 "transactions " : [ 7 { 8 "sender " : "b1c22f...5d8e " , 9 "message " : "Touch grass " , 10 "nonce " : 0 , 11 "signature " : " 8f23ab...45d1 " 12 } 13 ] , 14 "previous_hash " : " " , 15 "current_hash " : " " 16 } 17 ] 18 } Here, the payload list contains one block object (Node B’s proposal for block #2). Example — Values Broadcast (Multiple Blocks): If Node A, after collecting proposals from B (and maybe others), decides to broadcast the full set, it might send: 1 { 2 "type " : "values " , 3 "payload " : [ 4 { ... Block_A ... } , 5 { ... Block_B ... } , 6 { ... Block_C ... } 7 ] 8 } with each block from a different node. In all messages, JSON syntax must be strictly followed. That means: • Keys and string values in quotes " ". • Lists [ . . . ] and objects { . . . } with commas between elements. • No trailing commas. • True/False in JSON should be lowercase true/false if they appear. • Numeric values (nonce, index) appear as numbers (no quotes). In summary, design your networking code such that: • Before sending, you prepend the 2-byte length. • You then send the JSON text of the message. • When receiving, you first read 2 bytes to get N, then read N bytes to get the JSON message.
Department of Infrastructure Engineering GEOM90038 Advanced Imaging Lab Assignment 4: Parking occupancy detection from CCTV images Due for submission at 10:00 pm on Friday of Week 12 Note: This is an individual assignment. The task With the availability of the inexpensive cameras (and CCTV camera that are used for surveillance), computer vision methods have developed over the years for automatic object detection and classification. The state-of-the-art methods have achieved remarkable performance for object detection and classification, and include machine learning algorithms, such as convolutional neural networks. The aim of the assignment is to perform automatic parking detection from images, and to delineate the parking spaces automatically using machine learning. You will be provided a dataset that contains images from a publicly available dataset (PKLot) including images of occupied and unoccupied parking spaces. Using this dataset, an object detector (FasterRCNN model) will be trained to detect cars in images. Subsequently, the trained model will be used to detect cars and delineate parking spaces in another dataset (Barry street) automatically. Your task will be to perform. evaluation of the object detection algorithm, improve the parking space detections and subsequently plot the final parking space delineations. Figure 1. A screenshot of the parking occupancy detection using birds eye view images captured in Barry Street. Resources required You will use MATLAB 2020a or newer for the assignment. You can download it using the link (https://au.mathworks.com/downloads/web_downloads/download_release?release=R2020a). Please make sure to install all the packages (except the Simulink and related packages). Or alternatively you can use myUniApps (MATLAB 2020b). An .mlx file (to be opened using MATLAB) is provided in the LMS that will download the data and other required files for the operations. You can browse the data from the link (https://rmit.figshare.com/ndownloader/files/24753887). Please make sure to increase the java heap memory of MATLAB as described in the .mlx file. Another document (part of a book chapter) containing the background and the description of the process (including the dataset) is available on LMS. The procedure You will use the provided MATLAB live script. that has two tutorials in it, and a tool for visualising the dataset. In the first tutorial you will be able to fine-tune a pre-trained network (see book chapter for details) as a car classifier that can be used to classify the occupied and empty parking slots. In the second tutorial you will be able delineate the parking spaces automatically using spatio-temporal reasoning and using a density-based clustering algorithm. Thereafter, you need to write a MATLAB code for performing evaluation (calculating precision and recall), and to improve the delineation accuracy by using the statistics of the detections and assumptions of parking spaces. Following are the steps: 1. Visualise dataset: After downloading the .mlx live script, run the first section to download the dataset and unzip. The second section will load the bounding boxes of PKLot dataset and will annotate a few occupied parking spaces and will show an example occupied and empty parking slot. In the next section of the code, you will be able to visualise Barry street dataset and the parking spaces. 2. Create a car classifier: This step corresponds to Tutorial 1 of the live script. where a pre-trained classifier is fine-tuned with PKLot dataset containing images and labels. The input to the model is an image and the output is a decision, whether the parking slot is occupied or empty. Note, in its default, the provided code does not fine-tune the model, instead it loads the fine-tuned model for the experiments. To run it on-line one can set the option “train” to “true”. You can use the trained car classifier to test with Barry street dataset, and subsequently can estimate the accuracy, and some of the wrong estimations. 3. Delineate the parking spaces automatically and visualise: This step corresponds to Tutorial 2 of the live script. where you will perform. automatic parking slot delineation. A pre-trained object detector is fine-tuned with PKLot containing images and bounding boxes. The input to the model is an image of the whole parking space, and the output is the bounding boxes of the cars on the image. Similar to the previous step, you can set the “train” option to “true” to fine-tune the model on-line, note, however, this needs a GPU for training, otherwise it might take a very long time. You can use this trained car detector to detect cars for Barry Street images. However, you might realise, not all cars are detected. Also, not all parking spaces are occupied for the car detector to delineate them into parking spaces. Therefore, you can run the trained car detector on all the images of Barry Street dataset, and create a cluster around each parking space. Later, you can use a density-based cluster algorithm (DBScan) to find the location of the parking spaces. Important note: For the next steps, the code for performing the step must be included in the submission, otherwise this will lead to a penalty. 4. Perform. evaluation To perform the evaluation, you should write a MATLAB code to plot the precision vs recall of detections. You might realise that the value of precision is very poor, and in the next task you will improve it by using the statistics of the detections, and assumptions related to parking of a vehicle. The pseudocode: i. Adjust the variable “classifiedMean” to take into account the shift of 141 pixels along X axis, and 58 pixels on Y axis. ii. Create a blank table in the format that is accepted by the function “evaluateDetectionPrecision” and copy the box and scores of the transformed classifiedMean variable to the new table. iii. Calculate average precision, recall and precision, and subsequently plot the precision- recall curve as shown above. iv. The ground truth bounding boxes can be found in the file “GroundTruthBarryStreet.mat” . 5. Improve the delineation accuracy Write a MATLAB code and perform. the following steps: 5.1 Plot the statistics of the bounding boxes: Plot the bounding boxes for visualisation. You can rotate the bounding boxes to create an aspect ratio > 1, where the aspect ratio is defined by length/width of the bounding boxes. For the rotated bounding boxes, the X and Y coordinates, and the length and width of the bounding boxes should be interchanged. You can use the “rectangle” function of MATLAB for the plot. 5.2 Calculate the average length and width of the parking slots: Subsequently, calculate the mean length and width of the parking spaces for the whole parking area. 5.3 Post process the length of the parking slots based on assumptions: Assume that: • The length and width of the parking slots remain constant throughout the image. Therefore, use the average length and width for all the parking spaces. • The cars take 80% of the length of the parking space, and therefore, increase the length of the detections by 25% of the original length. The pseudocode: 1. Adjust the variable “classifiedMean” to take into account the shift of 141 pixels along X axis, and 58 pixels on Y axis. 2. For each bounding box in the transformed “classifiedMean” variable do: a. Check if aspect ratio is greater than 1, then do i. Calculate DeltaX as the difference between the bounding box width and the calculated average width ii. Calculate DeltaY as the difference between the bounding box length and the calculated average length of the car times a factor of 1.25 iii. To calculate the new X coordinate of the bounding box DeltaX/2 should be added iv. To calculate the new Y coordinate of the bounding box DeltaY/2 should be added v. Allocate the average updated length (1.25*average length) and average width for the new bounding boxes Else do: i. Calculate “DeltaX” as the difference between the bounding box length and the calculated average length of the car times a factor of 1.25 ii. Calculate “DeltaY" as the difference between the bounding box width and the calculated average width iii. To calculate the new X coordinate of the bounding box DeltaX/2 should be added iv. To calculate the new Y coordinate of the bounding box DeltaY/2 should be added v. Allocate the average updated length (1.25*average length) and average width for the new bounding boxes 5.4 Recalculate the precision vs recall and show final bounding boxes: Subsequent to post-processing, plot the precision and recall values again to show the improvement. Follow steps ii-iv of Task 4 to plot the precision-recall curve. Use function “rectangle” to visualise the new bounding boxes, and you might need to reverse the Y axis of MATLAB in order to generate the above plot. Submission Write an individual report outlining the process and your results. Include the following content: 1. Introduction: Covering the motivation of the assignment, the history of machine learning, computer vision and object detectors, and the evaluation matrices used for accuracy assessment. 2. Methods and results: Describe the processes you performed for completing each task, including visualise dataset, create a car detector, delineate the parking spaces automatically, perform evaluation and improve the accuracy performance. Include the screenshots of each process and any intermediate results. 3. Discussion: Describe about the accuracy evaluation of the parking slots, and show the improvement based on the made assumptions. Also, describe the challenges and the shortcomings of the performed method and propose scopes of improvements. 4. Conclusions: Provide a summary of your findings, highlighting the accuracy gain achieved before and after the pre-processing, and how it can be further improved. 5. Include the code snippets used for the calculations. Submit a digital version of your report, via LMS and in PDF format only. Marking rubric Introduction and object detection review 10% Proper description of the method 10% • Visualise dataset • Create a car detector • Delineate the parking spaces automatically Results • Precision vs Recall 10% • Plot the statistics of the bounding boxes 10% • Calculate the average length and width of the parking slots and plot them 10% • Post process the length of the parking slots on the 80% assumption 10% • Recalculate the precision vs recall and visualise bounding boxes 10% Discussion 10% Conclusions 10% Code 10%
Department Of Management Science MSCI 231 Introduction to Operations Management Module Outline 2024-25 Pre-requisites and limitations The module is available to students who have taken MSCI 100 / 101 or equivalent. It cannot be taken by students who have taken MSCI 102. Term Taught Michaelmas, weeks 1 - 10 Contact Time Approx. 30 hours (20 lectures and 10 seminars) Credits 15 Why take this module? Operations Management is a core managerial discipline for all kinds of operations - from private sector manufacturing through to public sector services. It is applicable to any of the organized processes that underpin the modern world: transportation, the generation of energy, retailing, the production of goods, the provision of medical and educational services and so on. Many areas of management have strong connections with operations management, so an understanding of its main principles is relevant to all students on business and management programmes. The nature of the subject A large part of operations management is analytical: structuring, measuring and reaching logical conclusions about operations problems - such as congestion, shortage, error and failure. Part of it is constructive: being able to design processes and put together plans that systematize, coordinate and improve work. The module reflects this combination, and includes both qualitative and quantitative methods. It is, however, grounded in practical issues and the experiences of organizations that provide case studies for the module. Learning outcomes By the end of the module students should be able to engage competently in the kind of problem- solving characteristic of operations management, including demonstrating an understanding of: • entities like supply networks; • use of mathematical models such as those needed to optimize inventory; • basic principles of lean production; • systematic techniques of control such as those needed for quality and project management; • sustainability in an Operations Management context; and • use of forecasting models in business. Teaching methods The module will be taught through a combination of lectures and seminars. There will be two lecture hours each week throughout the term, and one seminar hour each week starting in the second week of term. The seminars will be based on set problems that arise from the lecture material of the previous week. Attendance is required at all lectures and seminars and will be recorded by the University. Outline topics The following is a list of indicative topics that will be taught: 1. Operations as systems 2. Supply chain management 3. Inventory optimisation 4. Capacity analysis 5. Forecasting demand 6. ERP and lean principles 7. Project planning & control 8. Quality management 9. Sustainable Operations Assessment Students will be assessed through a 50% group coursework assessment (CWA) which will be set in week 10 (mid-December). The deadline for submission of the CWA will be announced later in the term, both in lectures during the module and also on Moodle. The remaining 50% will be assessed through an open book multiple choice (MCQ) examination during the summer exam period next year (2025). This will be an online timed exam. Guidance will be provided in week 10 on both coursework and examination. Workload The total workload for the module depends on the capacities of individual students, but in addition to the time needed for attendance at lectures and seminars, students are expected to engage with: • weekly discussion forums • short set readings which will be posted on Moodle • preparation and solution of case problems set for the seminars • coursework assignments • revision for the examination Reading and lecture notes Materials will be posted to Moodle at least 24 hours in advance of the lectures. These are not intended to be self-explanatory and attendance at the lectures is strongly advised. Model solutions or solution guides to each case dealt with in the seminars and problem classes will be posted on Moodle after the seminar class in the following week. The core text is Operations Management by Slack et al, published by Pearson. The current edition is the 10th, but earlier editions will also be mostly applicable. An e-version is available from the library: https://ebookcentral.proquest.com/lib/lancaster/detail.action?docID=6977559&pq-origsite=primo. Regular access to the core text will be needed for the set reading, the set cases, and for the further development of understanding of topics. Further Information The Management Science Department undergraduate coordinator is based in D25, Charles Carter. All announcements relevant to the module, as well as module materials, will be posted to Moodle. There is a departmental Moodle page with useful information for students on all MSCI courses at https://modules.lancaster.ac.uk/course/view.php?id=40859 Students are strongly advised to take note of warnings about late work, plagiarism and collusion. If unsure, please read the department's teaching code of practice, which can be found on the FAQ board. Contacting Linda Hendry - the module convenor As module convenor, I will be happy to answer any general queries about the module. The best time to catch me is at the end of a lecture. We will need to clear the room for the next group, but I will wait outside the lecture theatre after each lecture until I am sure none of you have waited to talk to me. I have also set up a general questions forum on the MSCI231 Moodle site - please check out this forum if you have a question. I prefer to use this rather than emails as many of you will have the same questions, and so I can post answers in the forum for the benefit of all students. I work part-time on Mondays, Tuesdays and Thursdays - and will aim to ensure that any questions on the forum are answered at the end of each of my working days. Of course, if your question is personal, you can email me. If so, please make sure you introduce yourself as being a student from this module, and I will answer as soon as possible on one of my working days. We hope that you enjoy the course and welcome feedback informally and via course reps.
CPT306 Individual Project Principles of Computer Games Design Coursework Assignment Specification 2024/25 Semester 2 Master’s Degree – Year 1 Coursework Assignment Number: 4 of 4 Method of Working: Individual Coursework Title: Creating a 3D Game Percentage (%) Weighting: 40% of the overall module marks Date and time of publication: 9:00 am on Tuesday, 15 April, Week 9 Date and time for submission: 11:59 pm on Sunday, 18 May, Week 13 General Instructions 1. One copy of this assignment should be handed via the module Learning Mall page at http://learningmall.xjtlu.edu.cn no later than the time and date shown above, unless an extension has been authorized by the module leader. 2. Before submission, each student must complete module coursework submission form obtainable from the module Learning Mall page. This assignment is being marked by student name and id, please ensure that you complete the correct coursework submission form. 3. Format of the coursework assignment submission: A ZIP file should be submitted via the Learning Mall module page, containing the deliverables outlined in the “What to Submit” section of the coursework assignment specification. 4. Use of unfair means: You are reminded of the University’s Code of Practice on the Use of Unfair Means and that the work you submit for assignment should contain no section copied in whole or in part from any other source unless where explicitly acknowledged by means of proper citation. 5. Late penalties: For work submitted late the penalty is loss of 5% marks per day. Work that is 5 or more days late will automatically be graded as FAIL, and no re-submission will be allowed. The story so far … An unknown entity forces the player into a gladiatorial arena. The player must overcome the gladiator creatures in order to earn the ability to escape the gladiator. Scenario, Player Movement, Release Skills, and Camera (30 points) A schematic diagram of the game scenario is shown in Figure 1. Figure 1: Concept of environments for survival game. The whole environment should achieve the following basic requirements (10 points): 1. Map and obstacle design . Map size: 35 x 35 units. . Obstacle Types: Obstacles contain two types, a base obstacle and a self-defined obstacle that is a patchwork stack of base obstacles. . Number of obstacles and height requirements: The map contains at least 15 base obstacles, each of which is a 1x1 unit square; the map contains at least 5 self-defined obstacles with a height of 2 units, i.e., a maximum of two 1x1 unit squares stacked vertically. . Obstacle Shapes: Obstacles may be shaped by combining one or more 1x1 unit squares into a custom shape, but may not exceed 2 units in height (i.e., a maximum of two 1x1 unit obstacles may be stacked vertically, no rules for horizontally). . Moveable Obstacles: Design at least 5 moveable obstacles to add a dynamic element. 2. Enemy design . There is only one enemy (Boss) in the game, excluding skill summoned enemies. . Boss Design: Boss can only move randomly within an 8x8 unit area and cannot leave the area.Boss movement can be as simple as randomly moving up and down or left and right. 3. Character Design . Boss Image: Boss image can be freely designed, but the size cannot exceed 4x4x4 units. . Hero Character: There can only be one Hero Character, and the image can be freely designed, but the size cannot exceed 1x1x1 units. 4. Skill Coin Designs . Maximum 3 Skill Coins: Each game can only have a maximum of 3 Skill Coins, which are used by players to unlock new skills. . How Skill Coins appear: Skill Coins appear randomly at certain locations on the map, and only one can appear at a time. . Skill Coin Image: Skill Coins can be freely designed, but cannot exceed 0.5x0.5 units in size. Note: If you fail to meet the above-mentioned requirements, it will affect the implementation and performance of subsequent functions, you will lose a significant number of points. The Hero’s movement control and release skills control should follow these methods (15 points): . Right mouse button clicks on the map (not on obstacles), the clicked place will appear a movement indicator (size not more than 1 * 1 unit), then the Hero will move to the indicator. Once the Hero moves to the indicator, the indicator will disappear. (5 points) . Left mouse button is used to release skills. Interaction methods can be freely designed. Potential interactions include, but are not limited to, the following: long-press to release a skill continuously, long-press and release to release a power-up skill, short-press to release a skill, and so on. The direction of the skill release and the target interaction can be freely designed based on the left mouse button interaction and the mouse position on the screen. (The Attack Indicator contains 5 points, with 5 points awarded for being easy to understand and creative; 2 points for each type of attack interaction, up to a maximum of 5 points) Camera of game scenes (5 points): . The center of the camera must always be aimed at the center of the map (Boss); . Press and hold Q and the camera will rotate clockwise around the map, press and hold E for counterclockwise. Skills Implementation (30 points) Both Boss and Hero in the game can release skills (they have the same selectable skills). The skills they carry in each game can be categorized into two types, attack skills and functional skills. Functional skills include defense skills and healing skills. Considering the different difficulties in implementing skills, we classify the difficulties into 5 levels as follows (when the implementation difficulty of a skill satisfies more than one level of difficulty determination, we will take the highest difficulty satisfied): Level 1. Release an entity that travels in a straight line (e.g. bullet). Level 2. Release a skill that moves along a certain trajectory (e.g., a parabola); Release a skill that requires energy storage; Release a skill that can add an effect to an entity directly (e.g. heals, invincibility, sustained damage). Level 3. Release an area-of-effect (AOE) skill; Release a skill that has an entity body and will self- destruct after a certain period of time. Level 4. Releases a skill that has an entity that can be attached to another entity (Boss, Hero, Floor, Obstacle, etc.) for a period of time. Eligible skills include, but are not limited to, mines, time-delayed bombs, etc. Level 5. The released skills have additional attributes and can interact with other objects. (For example, a flame skill can burn an enemy, a freeze skill can freeze an enemy, and an electric shock skill can paralyse an enemy for a short period of time.) The category of skills includes a total of 30 points. Achieving Level 5 directly earns 10 points. Achieving levels 1 and 2 awards 1 point. Achieving levels 3 and 4 awards 3 points. The highest cumulative score for Level 1-4 implementation is 20 points. It is important to note that the scores for Levels 1 through 4 above represent the base score only. If a demonstrated skill is not clear, scores will be deducted. Here is an example. If you implement an AOE skill without clearly indicating its range of effect, you may receive fewer than the full 3 points, potentially earning 2 or even 1 point instead. Additionally, there should not be the same skills appearing. Skills with the same effect but different numerical values are considered the same skill and should not be repeated. Skills Selection and Shopping System (15 points) The skill system in the game must meet the following requirements (6 points): . Hero and Boss both carry at least one attack skill and one functional skill; . Hero and Boss each carry a total of no more than 4 skills in the game; . In the game, players can switch the skill they want to release by using the four numeric buttons 1, 2, 3, and 4. In the main menu of the game, a Skill button needs to be included. After clicking to enter the Skill menu, players will be able to perform. the following operations (9 points): . Players can choose which skills to carry for the coming game (up to 4 skills); . Players can unlock the skills by using the Skill Coins obtain in the game; . The player's profile is automatically saved locally after exiting the game and can be automatically read when re-entering the game. User Interface and Gameplay (10 points) The user interface in the game needs to meet the following requirements (4 points): . Boss and Hero's health levels are clear; . Clearly display the currently selected skill in the game; . Clear display of skill cooldown time; . Display prompt messages correctly after Hero wins or is defeated. In addition, the default difficulty of the game needs to ensure that players can complete the game within 60-120 seconds (6 points). Training Ground (5 points) A training ground is accessible through the main menu of the game and consists of only two entities, the Hero and a Boss. In the training ground, the Boss does not have any skills but has an exceptionally high health level. Additionally, the player has the freedom to choose and release any skill for the Hero at any time, with the interaction/method of selection being freely designed. Note: As the training ground scenario can greatly assist you in testing your newly developed skills, it is strongly recommended to develop this scenario first. Functions that can help you hit high scores (10 points) Implementing the following requirements can help you achieve a perfect score for your assignment (4 points): 1. Has Background Music and Sounds Effect. 2. Contains a Difficulties Selection function. 3. Boss's skills are randomly selected, and Boss’s appearance is randomly generated. 4. Has a skill guidebook. In addition, an exquisite skill selection menu will help you earn up to 3 points, a well-made training ground can also contribute up to 3 points. Any requirements not mentioned will not help you earn more points. Implementation You have been asked to write your own code to implement the game using Unity and C#. The programming logic and environment for the game should be created by you. You can use materials from the web such as texture maps, models, bone animations. You may also use AI tools to help you implement simple features or to help you generate mapping material. Disallowed behaviors include, but are not limited to: extensive use of existing demos on the web, copying other people's assignments, assignments written by other people, game logic not written by yourself. Game specification After finishing the implementation of your game, you need to write a game description with document (.doc) file format. This document states the following information: . The version of Unity you used to develop the game. . The interactive methods of attacks in the game. . Skills Introduction. (Really Important) In this introduction, you need to provide the following information for each skill you have implemented: the difficulty level, specific numerical information, and the interaction method for releasing. . Whether you used AI tools, such as ChatGPT, to assist you in completing your assignment. If you used an AI tool, briefly state which AI tool you used and what problems you used the AI tool to help you solve. (Use AI tools will not influence your mark) If you have any special issues about the game, you can also write down in the game description. Some common questions Q1: Can obstacles be destroyed? A1: You can freely design. Q2: Can bullets / skill penetrate obstacles? A2: You can freely design. Q3: Can the Hero penetrate obstacles / Boss? A3: Under normal circumstances, it is not possible, but skills can be used to penetrate walls. Q4: Does the game have to end with either the Boss or the Player dying? A4: No. In order to fulfill the requirement of ending the game within 120s, you can add some game mechanics. These game mechanics include, but are not limited to, a 120 second countdown, where the game ends when the countdown reaches zero. Additionally, when the timer reaches the 90-second mark, the player's health gradually decreases, reaching zero by the 120-second mark. Furthermore, at the 90- second mark, the boss activates berserker skills, significantly increasing its damage output. It's important to make these effects as noticeable as possible in the game. If you have any further questions, please feel free to send an email. What to Submit Your ZIP file should include these files: . Submission form (.pdf). The submission form. should be properly completed with your signature. It is available on Learning Mall. Submission form. with incorrect information certainly will affect your marks, so carefully complete the submission form. The submission form. should be properly named as mentioned below. . The entire Unity project (folder). . Executable file of your game (.exe). Please build your game and include all the related built files in the submitted ZIP file. Note that, the target platform. should be Windows. . Game specification (.doc). This file is really important. . A skill demonstration video (.mp4). Display the release animation for each of your skills. Skill names and difficulty levels need to be marked. Up to 90 seconds. Please make sure all the required files can be opened and run properly on a Windows computer. Please use your First Name, Last Name and Student Number to name above mentioned files and the ZIP file—for example Xiaoming_Wang_999999 will be the name of the files module leader would be submitting, with 999999 being as his student number. Any submissions with improper or incomplete file names certainly will affect your marks, so carefully name your files.
Term Project Real Estate Finance STRUCTURING A HOTEL TRANSACTION FOR YOUR REAL ESTATE INVESTMENT VEHICLE Spring Term 2025 CONTEXT TO AVOID MISTAKES AND WASTED TIME AND EFFORTS, PLEASE READ AND REVIEW THIS DOCUMENT and FAQ IN DETAIL BEFORE STARTING YOUR WORK. Over the past decade, the low-interest-rate environment has driven investors away from cash and bonds, making higher-yielding assets like real estate more attractive. However, the post-pandemic capital market has been challenging and volatile. In the U.S., the annual inflation rate peaked at 8.0% in 2022, leading to a rising interest rate environment. This has put pressure on real estate values, as commercial real estate transactions are typically financed with significant amounts of debt. By 2023, the inflation rate began to stabilize, averaging 4.1%. As of February 2025, the year-on-year inflation rate stands at 2.8%. Concurrently, the U.S. economy faces political uncertainty, increased energy prices, geopolitical tensions, and extreme weather events. Despite these challenges, global tourism is expected to remain a secular growth industry in the long run. At the right purchase price, the current capital market environment may still offer attractive hotel real estate investment opportunities. The aim of this term project is to analyse and structure a hotel investment opportunity on behalf of your real estate investment vehicle. Your group represents a team of real estate investment analysts, whose primary task is to examine hotel investment opportunities for your real estate investment vehicle. Your firm is well-connected with the industry and regularly receives investment proposals from commercial real estate agents and market intermediaries such as CBRE, Colliers, or Cushman & Wakefield. In most cases, your team can reject proposals as inadequate for your investment vehicle, or too expensive. However, a number of recent opportunities have caught your interest. The chief investment officer has asked your team to investigate one of the investment opportunities in detail. In particular, your team is tasked with: 1) Identifying a suitable hotel investment opportunity for your investment vehicle. 2) Estimating the market value of your chosen hotel. 3) Performing a leveraged (before-tax) investment analysis and recommend an optimal transaction structure. Your team’s mission is to write up the findings of your investment analysis as a report that will serve as a basis for the investment committee to make a final decision regarding whether or not to acquire the hotel under consideration. Supporting Materials This term project handout will refer to the following resources on LMS: STR Reports: Forecast Report Profitability Report Trend Report Additional Materials: CoStar Market Reports CBRE Cap Rate Survey 2024 H2 News & Research Articles Once you dive into this term project, you will find an abundance of information at your disposal. One of the key challenges lies in your ability to critically filter and employ only the most relevant data that will effectively inform. your analysis. At the same time, you are strongly encouraged to broaden your research scope by integrating additional relevant information from credible sources to enhance the depth and breadth of your analysis. Expanding your research beyond the provided materials will not only enrich your project but also demonstrate a comprehensive understanding of the subject matter. To uphold academic integrity and ensure clarity in your report, make sure to properly cite all sources of information, including those listed above and any additional resources you consult, following the APA reference style. Proper citation is essential for enabling readers to verify the information presented and comprehend the foundation of your analysis. Furthermore, both the written report and accompanying Excel file must be presented in a manner that is easily accessible and understandable to someone who may not be familiar with the intricacies of your project, such as your lecturer or a busy member of an investment committee. This requires careful attention to the organization, clarity, and presentation of your work to ensure it can be efficiently comprehended by readers with limited time. By adhering to these guidelines and expectations, you will not only fulfill the academic requirements of the project but also produce a professional-quality analysis capable of realistically informing investment decisions in a professional setting. 1.Hotel Choice When analysing potential investment opportunities, it is essential to consider the characteristics of your investment vehicle. Specifically, what risk-return profile is your investment vehicle targeting? This profile will not only influence whether you invest in core versus opportunistic opportunities but also affect the optimal amount of financial leverage to use when structuring the transaction with equity and debt. For this term project, your first task is to choose a hotel for your investment vehicle: Appendix 1 shows which group has which real estate investment vehicle allocated. Appendix 2 shows the list of hotels you may choose. Several hotels on this list could be suitable choices for your investment vehicle, with no definitive “wrong” options. The primary objective of this initial task is to demonstrate your understanding of your vehicle’s investment background and typical targets. Additionally, this task allows you the flexibility to select which hotel and sub-market you wish to analyse in detail. The aim is not to conduct a systematic or extensive preliminary analysis of all hotels. You should spend only a few hours making this selection. However, you will need to write one or two paragraphs in the executive summary of your report explaining why your chosen hotel is a suitable investment candidate, considering your investment vehicle’s background. 2.Market Valuation 2.1 Cash Flows This task involves estimating the market value of your chosen hotel as of June 1st, 2025, using a 10- year discounted cash flow (DCF) analysis. The first step in the market valuation is to provide an overview of your hotel and its location. Describe the facilities, size, and scale of the hotel, the strength of the brand, and the micro-location. Analyse supply and demand in the local hotel market by considering the strength of the local economy and the area’s attractiveness to tourists. The STR trend report will be helpful in understanding the evolution of occupancy, supply, and demand in the market. This preliminary analysis will serve as a basis for estimating expected future cash flows and the property discount rate in the market valuation. You are encouraged to create graphics or tables (e.g., a map showing the micro-location of the hotel and its major competitors). Overall, the description of the hotel and its location should not exceed one page of text (excluding graphics or tables). The next step is to estimate the future cash flows of your hotel for the next ten years. Determine the average daily rate (ADR) of your hotel using current room price information available online. If the hotel has multiple room categories, calculate a weighted average ADR that reflects the number of rooms in each category. Additionally, consider seasonal variations in room prices to ensure your ADR is representative of the entire year. If you cannot find room rates for the whole year, use data from the STR Trend Report to extrapolate the missing figures. Note that hotel prices per room shown on the internet represent the so-called "rack rate." According to industry insiders, the ADR received by the hotels tends to be ~ 15% lower than the rack rate. Therefore, adjust your online prices downward accordingly. Another important factor to consider when estimating the ADR for your hotel is the value-added tax (VAT). In many countries, the room rate advertised online includes VAT, which is a tax paid by the guest to the government. However, this tax is not a revenue stream for the hotel and should be excluded from the calculation of the ADR. Therefore, it is important to divide the rack rate by (1+VAT rate) to obtain the pre-tax room rate, which is used to estimate the hotel's future cash flows. To calculate the ADR growth rate (for the stabilised phase), utilise the four-factor formula introduced in the lecture. This involves considering the historical ADR growth in the market, the depreciation rate, and both past and anticipated future inflation rates. Should your analysis determine that the market is currently undergoing a transition phase, it's crucial to model the ADR explicitly for the duration of this phase. The ADR growth rate derived from the four-factor formula should then be applied starting from the first-year post-stabilisation. Use the STR Trend Report to calculate the historical ADR growth rate in the market based on CAGR. Hint: Ideally, use a historical data period as long as possible to cover several market cycles. To estimate the depreciation rate, you should ideally use a simple linear regression model to determine a submarket-specific rate. The STR Profitability Report for your hotel’s scale includes an MS Excel sheet called “Participation Report,” which can serve as a starting point for building a relevant comp set. Gather data on the current ADR (sourced online) and the age of hotels from your selected comp set. Ensure your final depreciation rate estimate is realistic, which may involve eliminating outliers (e.g., an old hotel with a very high ADR due to its location). Your estimate may become more accurate as you increase the sample size by including relevant hotels not listed in the STR participation report. Ultimately, you may still choose to adjust your estimate manually. In the lecture, we discussed that the depreciation rate typically ranges from -0.5% to -1.5% per year. If your regression estimate is 0% (or even positive), consider it an indication that the depreciation rate in this market is likely at the lower end of the range (-0.5%). Use online resources to find the historical inflation rate for the same period. Develop a realistic estimate for the future inflation rate. In the lecture, we typically represented the future inflation rate as a single figure, resulting in one future ADR growth rate. However, given the current high inflationary environment, you might consider forecasting the inflation rate explicitly for each of the next 10 years, or at least until it stabilizes. Consequently, the ADR growth rate would vary for each year or phase Hint: Always carefully consider whether your inputs and estimations are realistic. Consistently sense-check inputs and results, as this applies to all aspects of your analysis. Avoid inputting your result from the four-factor formula into the DCF if you find the estimate unrealistic. Use your analytical judgment to determine if adjustments are necessary. Such adjustments may be warranted based on insights gained from analysing future supply and demand. To estimate a realistic current occupancy rate, start with the market-level occupancy rate provided in the STR Reports. Carefully assess whether your hotel’s occupancy rate might differ from the market average. For instance, hotel-specific occupancy rates may need to reflect the unique characteristics of the hotel’s submarket, micro-location, or other peculiarities. Ensure you account for any cycles and seasonality appropriately. When you forecast changes in the occupancy rate, be cautious not to end up with unrealistically high or low estimates for the final year. To estimate other segment revenues (e.g., F&B and Spa), use the benchmarks provided in the STR Profitability Report. In a real-world scenario, you would have access to the hotel’s historical revenues and expenses. For this term project, we will use the STR benchmarks as an approximation. Keep in mind that the figures in the STR Profitability Report represent the market average. State your assumptions for other revenue and cost items. Explain your choice of starting year values (e.g., as a percentage of revenue or in absolute terms) and their growth over time (e.g., in line with the expected inflation rate). The challenge is to make the market valuation as specific to your hotel as possible. While market averages can often serve as reasonable proxies, your hotel may differ from the market average in many cases. Customize the market valuation as much as possible. For instance, eliminate revenue and cost items for “F&B” if your hotel does not have a restaurant or bar. Clearly state the reasons for any upward or downward adjustments. If you find no reason to deviate from the STR benchmarks, mention this as well, providing a brief explanation. Critically analyse whether the reserve for capital replacement in the STR Profitability Report sufficiently approximates all actual future Capex requirements. Adjust it if necessary. Addtionally, evaluate if franchise, management, and incentive fees are appropriate for your hotel. Use the direct-capitalisation approach to estimate the expected sales price at the end of the 10th year. Start with the proxies provided in the lecture slides for estimating buyer and seller transaction costs, but also consider finding country-, sector-, or even city-specific proxies that are more realistic for the particular circumstances. Avoid being misled by data artifacts. For instance, if you estimate an extremely high or low depreciation rate, interpret it as an indication that the depreciation rate in this market is at the upper or lower end of a reasonable range, as discussed in the lecture. When encountering conflicting data from different sources, use your discretion to select the most realistic option and explain your choice. 2.2 Property Discount Rate Discount the expected future cash flows using a property discount rate that appropriately reflects the hotel's risk profile. Refer to market reports (such as the one mentioned below) and apply the cap rate approach taught in the REF lecture to estimate the property discount rate. Finding current market hotel cap rates can be complex, depending on your market. You are encouraged to seek out the best available information, with CoStar reports being particularly useful. Consider making further property-specific adjustments to account for factors such as micro-location, competitive situation, and other hotel-specific risks (e.g., building-related issues). Aim to account for these differences to the best of your knowledge and justify your reasoning. Consider whether current cap rates are in equilibrium or unusually high (low) due to the hotel market being in a crisis (boom). If you suspect this is the case, a complementary estimate of an appropriate property discount rate based on the risk-premium approach may be helpful. Ensure that your estimated cap rate and forecasted future growth rate are realistic. Emphasise making strong arguments to support your rationale. When estimating the growth component in the cap rate approach, you may use your result from the ADR growth rate analysis based on the 4-factor-formula. Finish your market valuation by stating the rounded market value of your hotel (rounded to full 100,000 $.) The following investment analysis requires a specific assumption for the purchase price. While the market value estimates the expected sales price under normal conditions, the actual transaction price is always the result of negotiations between the buyer and seller. For instance, a seller under financial distress may sell under time pressure, negatively impacting the sales price. Conversely, a buyer may identify defects in the building’s physical structure. A skillful negotiator can leverage such shortcomings to secure a lower purchase price than would otherwise be justified. For the investment analysis in the next step, assume that after a long, strenuous, but successful negotiation process, your investment vehicle (the buyer) has agreed with the seller on a transaction price 10% lower than your estimated market value. This discount was possible because the previous owner had excessive financial leverage and was forced to sell to improve liquidity and avoid bankruptcy. Overall, the market valuation section of your term project should not exceed four pages, excluding tables and figures. 3.Investment Analysis 3.1 Setting The primary objective of the term project is to assess whether the potential hotel acquisition target meets the risk-return requirements of your investment vehicle. This entails ensuring that the hotel has the potential to at least match, and ideally exceed, the minimum total return target after accounting for financial leverage. Additionally, the hotel should align with the risk tolerance of the typical investors in your investment vehicle. The investment analysis aims to determine whether the investment is financially advantageous and to propose a transaction structure that best aligns with your investment vehicle’s risk-return preferences. Consider the following minimum total return targets for the different investment vehicles. Simultaneously, ensure that the maximum loan-to-value (LTV) ratios are not exceeded to maintain controlled investment risk: For more information on the three investment vehicles, refer to the lecture slides "Real Estate as an Asset Class" and the short article "The Major Real Estate Investment Vehicles" on LMS. This part of the term project aims to structure the investment transaction using an optimal mix of equity and debt financing. Examine the following three alternative transaction structures with respect to the LTV ratio: Assume that the mortgage loan is an interest-only loan, fully repaid by the end of the final year of your investment analysis. For the interest rate assumption, which depends on the LTV ratio, refer to the instructions provided below. 3.2 Mortgage Loan Interest Rates as a Function of LTV When evaluating the optimal financial structure, consider the increased risk associated with higher financial leverage. As detailed in the lecture “The Effect of Leverage on Risk & Return,” both the expected return and risk for the equity investor rise with an increasing LTV ratio. Additionally, the interest rate on mortgage loans tends to increase with the LTV ratio, reflecting the heightened risk for the bank as financial leverage escalates. Therefore, as part of the investment analysis, it is essential to determine the cost of debt—or the mortgage interest rate for hotel loans—across various LTV scenarios. The HVS Report on hotel interest rates, published on May 30, 2024, provided estimates for a range of potential hotel interest rates in the coming years as follows: Given the uncertainties surrounding borrowing costs for hotel investments, and for the purposes of this project, refer to the table below for the borrowing rate to use in your hotel analysis. 3.3 Leveraged Investment Analysis Before Taxes The final step is to conduct the leveraged investment analysis before taxes. Calculate the leveraged cash flows for the three LTV scenarios (see above) depending on your investment vehicle. Afterwards, analyse the following three measures for each LTV scenario: Internal rate of return (IRR) Modified internal rate of return (M-IRR) Net present value (NPV) Which investment decision tool prefers which LTV scenario? Explain any potential differences in the rankings and come up with a final recommendation. The last paragraph of the investment analysis should conclude with a summary of the investment case. Provide a recommendation regarding whether or not to pursue the transaction and the optimal transaction structure. Overall, the investment analysis part of your case study should not exceed three pages (excluding figures and tables). 4.Deliverables 4.1 Professionally Formatted Report (PDF) The report must contain the following sections: 1. Executive Summary (about 1 page): A concise summary of the context, your findings, and your recommendation. o Brief explanation of why you chose the hotel and how it matches your investment vehicle. o Market value estimate from the DCF approach. o Optimal LTV structure and the corresponding interest rate. o A brief explanation of the final recommendation for the deal structure. 2. Market Valuation (about 3-4 pages): How did you develop your assumptions regarding growth rates, cash flows, cap rates, discount rates, capital reserves, etc.? 3. Investment Analysis (about 2-3 pages): Interpret the financial analysis results. Explain your recommended deal structure. Why did you choose this specific loan scenario, and how does it fit into the investment strategy? Requirements: Strictly within seven pages of text, i.e., excluding tables, figures, and the cover page. The more concise your report, the better. APA referencing is mandatory in the report but not in MS Excel. The report must be formatted in Times New Roman, with line spacing 1.5, a font size of 11, and standard top and bottom margins (circa 1 inch). Any violations of these rules result in a penalty. 4.2 Financial Model (MS Excel) The Excel file must contain the following tabs in the given sequence. Additional analyses could be added in the latter tabs: 1. Introduction: A short description of the MS Excel file. A layperson reviewing the file should be able to comprehend the nature of the contents included in the file. 100-150 words are ideal. 2. Preliminary Analysis: This part may contain the analysis of the cap rate and growth rate. The sheet may also include tables on other statistics such as ADRs, pipeline reports, occupancy ratios, or measures to assess the risk of the market. No extensive explanation is needed. 3. Assumptions: All assumptions and sources used in the DCF valuation method must be specified. The remaining tabs must use cell-referencing in calling these assumptions. The absence of cell-referencing may affect your grading if we cannot efficiently check your calculations. 4. Market Valuation: A complete DCF analysis. The final value should be provided in an easy-to-spot, specially formatted cell. 5. Investment Valuation: This sheet should contain the leveraged cash flows for three different LTV scenarios, as well as the calculation of the IRR, M-IRR, and NPV for all three scenarios. You can opt to show the different LTV scenarios in separate tabs. 10
Course: Ontario Secondary School Literacy Course Course Code: OLC4O Department: English Course Developer: Canada eSchool Development Date: 2012 Revision Dates: 2022 Grade: 12 Course Type: Open Credit Value: 1.0 Prerequisite / Eligibility: Students who have been eligible to write the OSSLT at least twice and who have been unsuccessful at least once are eligible to take the course. (Students who have already met the literacy requirement for graduation may be eligible to take the course under special circumstances, at the discretion of the principal.) (Ministry of Education [MOE], 2003, p. 19). Curriculum Policy Reference: The Ontario Curriculum, English: The Ontario Secondary School Literacy Course (OSSLC), Grade 12, 2003 Course Description This course is designed to help students acquire and demonstrate the cross-curricular literacy skills that are evaluated by the Ontario Secondary School Literacy Test (OSSLT). Students who complete the course successfully will meet the provincial literacy requirement for graduation. Students will read a variety of informational, narrative, and graphic texts and will produce a variety of forms of writing, including summaries, information paragraphs, opinion pieces, and news reports. Students will also maintain and manage a portfolio containing a record of their reading experiences and samples of their writing (MOE, 2003, p.19). Overall Expectations Building Reading Skills By the end of this course, students will: • demonstrate the ability to read and respond to a variety of texts; • demonstrate understanding of the organizational structure and features of a variety of informational, narrative, and graphic texts, including information paragraphs, opinion pieces, textbooks, newspaper reports and magazine stories, and short fiction; • demonstrate understanding of the content and meaning of informational, narrative, and graphic texts that they have read using a variety of reading strategies; • use a variety of strategies to understand unfamiliar and specialized words and expressions in informational, narrative, and graphic texts. Building Writing Skills By the end of this course, students will: • demonstrate the ability to use the writing process by generating and organizing ideas and producing first drafts, revised drafts, and final polished pieces to complete a variety of writing tasks; • use knowledge of writing forms, and of the connections between form, audience, and purpose, to write summaries, information paragraphs, opinion pieces (i.e., series of paragraphs expressing an opinion), news reports, and personal reflections, incorporating graphic elements where necessary and appropriate. Understanding and Assessing Growth in Literacy By the end of this course, students will: • demonstrate understanding of the importance of communication skills in their everyday lives - at school, at work, and at home; • demonstrate understanding of their own roles and responsibilities in the learning process; • demonstrate understanding of the reading and writing processes and of the role of reading and writing in learning; • demonstrate understanding of their own growth in literacy during the course. (MOE, 2003, pp. 20 - 26) Outline of Course Content Title Time Module 1 Personal Choices 29.0 Hours Module 2 Community Voices 24.0 Hours Module 3 Community Action 26.5 Hours Module 4 Leading into the Future 20.0 Hours Independent Study Unit: Literacy Journal Portfolio 10.0 Hours OLC4O Final Examination 3.0 Hours Total Course Hours 112.5 Hours Resources Peterkin, P., ed. Literacy Power OSSLC. Toronto: Gage Learning, 2004. Topic specific websites and articles are listed when required. Teaching/Learning Strategies The strategies used are varied to meet the needs and the range of learning styles encountered in any group of students and include the following: • Problem Solving • Reports • Independent Study • Guided Internet Search • Group Discussions • Textbook Use • Email • Independent Reading • Direct Instruction • Instant Messaging/Chat Room • Research • Verbal Discussion Strategies for Assessment and Evaluation of Student Performance Assessment is a systematic process of collecting information or evidence from a variety of sources that accurately reflect how well a student is achieving the curriculum expectations in a subject or course. The primary purpose of assessment is to improve student learning. Assessment for the purpose of improving student learning is seen as both assessment for learning and assessment as learning. In assessment for learning, teachers provide students with descriptive feedback and coaching for improvement. Assessment as learning is achieved by helping all students develop as autonomous learners who are able to set individual goals, monitor their own progress and reflect on their thinking and learning. Information gathered through assessment helps teachers to determine a student’s strengths and weaknesses and serves to guide teacher feedback. (MOE, 2010, Ch. 4). Evaluation refers to the process of judging the quality of student work on the basis of established criteria. Student evaluations are an assessment of learning focusing solely on a student’s achievement of the overall curriculum expectations in his or her course. Student marks are not averaged; instead, a student’s final mark is based on their most consistent effort with special emphasis on their most recent work. CATEGORIES OF KNOWLEDGE AND SKILLS Strategy Purpose Who Assessment tool Ongoing communications (between teacher and student) assessment for learning assessment as learning Self, Teacher Anecdotal records Asynchronous Discussions assessment for learning Self, Peer, Rubric assessment as learning Teacher Online Activities (questions, reflections, case studies, self-assessment quizzes, and/or research) assessment for learning assessment as learning Self, Teacher Rubric Asynchronous Discussions assessment of learning Teacher Rubric Online Activities (assignments) assessment of learning Teacher Rubric Unit tests and quizzes assessment of learning Teacher Marking Scheme Independent study unit assessment of learning Teacher Rubric Final exam assessment of learning Teacher Marking Scheme Considerations for Program Planning: Teaching Approaches Teachers will provide a wide range of activities and assignments that encourage understanding and mastery of the course curriculum. To make their programs interesting and relevant, they will help students to relate and apply the knowledge and skills gained in this course to issues and problems in the real world. Planning English Programs for Exceptional Students In planning courses for exceptional students, eSchool teachers will examine both the curriculum expectations for the course and the needs of the individual student to determine which of the following options is appropriate for the student: • no accommodations or modifications; or • accommodations only; or • modified expectations, with the possibility of accommodations If the student requires either accommodations or modified expectations, or both, the relevant Information must be recorded in his or her Individual Education Plan (IEP). Students Requiring Accommodations Only. With the aid of accommodations alone, some exceptional students are able to participate in the regular course curriculum and to demonstrate learning independently. (Accommodations do not alter the provincial curriculum expectations for the course.) The accommodations required to facilitate the student’s learning must be identified in his or her IEP. If a student requires “accommodations only”, assessment and evaluation of his or her achievement will be based on the appropriate course curriculum expectations and the achievement levels outlined in this document. English As a Second Language and English Literacy Development (ESL/ELD) Young people whose first language is not English enter Ontario secondary schools with diverse linguistic and cultural backgrounds. Teachers will incorporate appropriate strategies for instruction and assessment to facilitate the success of the ESL and ELD students in their online course. These strategies include: 1. modification of some or all of the course expectations, based on the student’s level of English proficiency; 2. use of technology to provide a variety of instructional strategies 3. use of a variety of learning resources (e.g., visual material, simplified text, bilingual dictionaries, and culturally diverse materials); 4. use of assessment accommodations (e.g., granting of extra time, use of oral interviews) Antidiscrimination Education Learning activities and resources used to implement the course curriculum are inclusive in nature, reflecting diverse points of view and experiences. They enable students to become more sensitive to the experiences and perceptions of others. The Role of Technology Information and communications technology (ICT) provides a range of tools that can significantly extend and enrich teachers’ instructional strategies and support students’ learning. Teachers will use ICT tools and resources both for instruction and for the design of curriculum to meet diverse student needs. ICT may be used to connect students to other schools, locally and abroad, and to bring the global community into the local online classroom. Through internet websites and CD-ROM technology, students can now access primary resources held in museums, libraries, archives, and public institutions across the country and around the world. ICT resources allow secondary students to conduct more far-ranging and authentic research than ever before. Applications such as databases, spreadsheets, word processors, and presentation software can be used to support various methods of inquiry. The technology also makes possible simulations of complex systems that are useful for problem-solving purposes or when field studies on a particular topic are not feasible.
MATH2022: Linear and Abstract Algebra Semester 1, 2025 Assignment 2 1. Using the definition, carefully show that each of the following are subspaces of the given vector space. (a) W = {(a1, a2, . . . , an) ∈ Fn : a1 + a2 + · · · + an = 0} ⊆ Fn (b) Z = {f ∈ RR : f(a) = 0 for a fixed number a ∈ R} ⊆ RR (c) For L : Fn → Fm a linear map, the set Im(L) := { ∈ Fm : there exists some ∈ Fn such that = L()} ⊆ Fm 2. Give a proof for the following (unrelated!) statements about linear dependence and independence. (a) Let V be a vector space over a field F, and let be a collection of vectors. Suppose that Y ⊂ X is a nonempty subcollection of these vectors, and we know that Y is a linearly dependent in V . Show that the collection X is necessarily linearly dependent. (b) Let V be a vector space over a field F, and let be a basis for V . Let ∈ V be a nonzero vector, and suppose it has coordinate vector with µ1 ≠ 0. Prove that the set is also a basis of V . Remark: There is nothing special about µ1 here! What we’re really saying is that there is at least one of the basis vectors in B that we can swap out for a nonzero vector and still get a basis! I’ve just phrased it as above for clarity. 3. Let P2(R) be the vector space of all two-variable polynomials with real coefficients, in the variables x and y. For example, we have elements like (a) We define two maps where and are the partial derivatives with respect to the variables x and y, respectively. Show that E and F are both linear maps. Note: If you are not quite familiar with partial derivatives, fear not! They behave much like regular derivatives. Check out the many resources online or from friends about how partial derivatives work. The key idea is that “treats y like a constant,” and similarly for . So for example, if p(x, y) = x4 − 3x2y3 + 2y − 7, then taking the derivative of the x-variables and pretending y is a constant, we get and similarly treating y like a variable and x like a constant, we get (b) Let be the subspace of two variable polynomials of total degree three; that is, is the subspace spanned by the basis Show that in fact both E and F map P 2 3 (R) → P 2 3 (R). Do this by showing that E and F send each of the basis elements to elements in . (c) For a vector space V and a linear map T : V → V , we say that a vector ∈ V is an eigenvector of T if T() = λ for some scalar λ. Define the linear map H : → as the difference of the compositions H := E ◦ F − F ◦ E. Show that the given basis of are all eigenvectors of H, and find their eigen-values.
MAST10006 Calculus 2, Semester 1, 2025 Assignment 6 Due by: 12pm (midday) Monday 19 May 2025 ❼ Answer all questions. Of these questions, one will be chosen for marking. ❼ Submit your assignment in Canvas LMS as a single PDF file before the deadline above. ❼ Marks may be awarded for: ➒ Correct use of appropriate mathematical techniques. ➒ Accuracy and validity of any calculations or algebraic manipulations. ➒ Clear justification or explanation of techniques and rules used. ➒ Clear communication of mathematical ideas through diagrams. ➒ Use of correct mathematical notation and terminology. ❼ You must use methods taught in MAST10006 Calculus 2 to solve the assignment questions. ❼ Give any numerical answers as exact values. The Gompertz model for a population is where k and a are positive constants, and p(t) is the population size at time t. The Gompertz model was first studied by Benjamin Gompertz in 1825. Gompertz was an actuary who used the model to investigate human life expectancy. It has since been used by researchers from various fields to model the growth of different ‘populations’, including by market researchers to model the uptake of a new product amongst consumers, and by oncologists to model the growth of cancer tumours. Let Then the Gompertz model can be expressed as Question 1. You may use the following facts about f(p) in this question: ❼ f(p) = 0, and ❼ = 0 only when p = , and this point is a local maximum of f (where e = 2.71828 . . . is the usual exponential base). (a) Find the equilibrium solution(s) of the Gompertz model in terms of k and a, or show that it does not have any. (b) Draw a phase plot for the Gompertz model. Label all important features with their values, including any local maxima or minima. (c) Sketch the family of solutions for the Gompertz model. (d) Describe the long-term consequences for the population predicted by this model. (e) Suppose that initially the population is p(0) = . What is: (i) the initial rate of growth of the population? (ii) the maximum rate of growth of the population? Question 2. Suppose that a population is modelled by the Gompertz model with k = 2 and a = 10, and that initially the population has size p(0) = 0.1. Find p(t) in terms of t, and find how long it takes until the population reaches size p = 5. Question 3. Prove the two facts given in question 1: (a) Show that f(p) = 0 (b) Show that = 0 only when p = , and that p = is a local maximum of f.
Spring 2025 CS 14 Individual Project CSCI 14 Course Project, Baseball, 200 Points Draft on 5/2/25, Final Expected on May 6 or May 7. (This doc will be slightly revised after student feedback to include clarifications, improvements, and more defined grading criteria. Assignment to be fully launched on May 7.) Array of Struct with file input and output & Functions The local baseball team is computerizing its records. You were asked to write a program that computes batting averages and other statistics. There are 20 players on the team, identified by the numbers 1 through 20. You will have two input files: (find them in Canvas). You need to test run your program using those 2 files (1) A Roster File – A file contains 21 lines, first line is an integer indicating the season. The rest is in the format “last name first name” (separated by a space) for each player, as follows: 2022 Smith Sam Jones Bob Miller Miles (20 names for 20 players) and etc….. (The players’ ID are numbered 1-20 in the order listed in this file.) Noted: this file contains no player numbers, just their names. Assign the player ID while reading (2) A Statistic File – A separate file contains the game-by game statistics for the players as a “player ID number, hits, walks, outs” 3 4 3 0 9 7 3 0 11 4 2 2 9 1 1 1 3 5 2 8 and etc……. Here is an example: 20 2 1 1 The example above indicates that during a particular game in the season, player number 20 was at bat four times and made 2 hits, 1 walk, and 1 out. There might be several lines in the file for a player. (a player can play more than one game in a season) The file is organized by the batting line-ups (it is not sorted by player number). Each player’s batting average is computed by adding the player’s total number of hits dividing by the total number of times at bat. A walk does not count as either a hit or a time at bat when the batting average is being calculated. When your statistics file contains no data, print message to the output file saying, “The [2025] baseball season was cancelled” and end the execution. [2025] is stored in a variable that was read from the Roster file. Create an array of PlayerInfo struct to hold the players’ data. The array is local to main function and passes to other functions as argument. struct PlayerInfo { int PlayerID; string LastName; //struct declared in global area (above main & function prototypes) string FirstName; int Hits, Walks, Outs; double Batting, OnBase; }; PlayerInfo Player[20]; //array declared inside main function and pass it to sub functions. GRADING POINTS will be given based largely on the BASE requirements and the PROGRAM EXECUTION requirements. BASE Requirements: (All Functions must Be below the main function!!!) Don’t use the following in your program at all: no vectors, no stacks, no multi-dimensional arrays, no parallel arrays or any other C++ coding techniques not covered by the Course. Significant number of points may be taken off No global variables allowed. Global area can only have function prototypes, named constants, and struct declaration. (your program must follow this logic) Functions must be called in this order inside the main function. 1. In the main function, you will call a function GetPlayer to read from Roster file and store the player names and id into the array and return the year for the season. int GetPlayer (PlayerInfo [ ] ); // function prototype (Be careful! The players’ identification numbers are 1 through 20, but C++ array index starts at 0). 2. Call a PrintRoster function to send the roster of the team with player ID in order to an output file 3. Call a SortPlayer function to sort players by last name. (i.e. Use Selection or Bubble Sort) NOTE: The (index + 1) of the array is no longer the player’s ID after sorting. 4. Then call a GetStatistics function to read the statistics file, updating each player’s statistics of hits, walks and outs. NOTE: – you need to keep the player identification numbers with the player names and the data in the struct. When you read a player’s game line, look up the player in the array of structs before you adding up the hits, walks and outs. Hints: use Linear (or Binary Search)Search to look up the player id, before adding hits, walks, and outs to that player. if (ID = = Player[index].PlayerID) { Player[index].Hits += H; Player[index].Walks += W; Player[index].Outs += O; } 5. A function to calculate each player’s batting average and on base average and store them into the array. (remember to do casting in order to retain the factional part) batting average = double (hits) / (hits + outs) on base average = double (hits + walks) / (hits + walks + outs) 6. A function to Send the players’ statistic to the output file including their Batting Average, On Base Average, in DL list, the best hitter, the worst hitter, the best base runner (based on OBA) and the worst base runner. Note: A player is in the DL (disabled list) if he didn’t play at all in the entire season. Sample of an output file: Baseball Season 2025 -------------------------------- Smith Sam Player 1 Jones Bob Player 2 Miller Miles Player 3 and etc…… Rutter Jack Player 20 Player Statistic with their total hits, total walks, total outs, batting and on base average for season 2025 Player Number Hits Walks Outs Batting Average On Base Average In Disable List ------------------------------------------------------------------------------------------------------------------------------------------- Apbee Cory 5 5 3 5 0.50 0.62 Buzz Bee 8 3 0 0 1.00 1.00 Carter John 4 Yes Crane David 19 3 4 4 0.43 0.64 and etc…… Give the name(s) and batting average of the best hitter (i.e. highest batting average). Give the name(s) and on-base average of the best base runner Give the name(s) and batting average of the worst hitter Give the name(s) and on-base average of the worst base runner If there is a tie for best or worst, Print all the players’ names. Just an Example: (might not be true for your input file) The best hitter with batting average of 0.68: Jones The worst hitter with batting average of 0.04: Miller, Smith 2) Program Execution Requirements: Possible adjustments to grade Points included in 200 points, fully or partially deducted if not done or done partially. You are required to use the provided dataset named statistic.txt and roster.txt and not only your own custom dataset, - 20 pts Attach documentation of a screenshot of program compilation and execution, your choice of dataset with date/time included, - 20 pts Attach the output file with the results of the program - 20 pts. If you only print it to the screen you can lose up to 20 pts If you provide any file names that have file paths (or are not just one filename), your program will be considered non-executable and you will receive a maximum of ONLY 100 points, with other possible deductions based on compilation and reviewing your code. Points possible that can be added to 200 points (Considered extra credit, where this can help offset other points off or allow to get more than 20 points) – there is some risk to choosing these options. If you don’t try this, it is ok, but you not received extra credit points. If you try and it is not done well, you can receive minimal extra credit points or worse yet it can impact points on the other grading criteria. Datasets Options: Encourages you to use the file naming convention or file name entry 1) Use the provided datasets and one other custom dataset, specify the file names in two different .cpp files a. Provided dataset file names are: statisticxx1.txt, rosterxx1.txt1 and resultsxx1 for the first .cpp file, +10 pts b. This can only be done if you choose to do a test run also with a Custom dataset file names are: statisticxx1.txt, rosterxx2.txt2 and resultsxx2 for the first .cpp file, +10 pts OR 2) Allows the user to enter the statistics, experts and output file names. +15 pts to +25 pt. You must attach all three files after the completion of each test run (for one test for the provided dataset +15 pts and two tests if you also run it for a custom dataset. +25 pts No need to test your program more than one test for the provided dataset and one test for the custom dataset. You will not receive additional extra credit points.
Assignment 3 FINC6024 Total Marks: 10 1) [2.5 Marks] Case 3: Executive Summary of the Case 3 Resolution (max 200 words). Provide an executive summary of your team's case 3 resolution. Your executive summary should focus on the macro-level reasoning and recommendations. It may help you to imagine that this ES will be the first thing (on your resolution) read by the CEO. 2) [5 Marks] Optimal Capital Structure of A-REITs & Stapled Securities Should A-REITs be highly leveraged, or lowly leveraged? Describe the optimal capital structure for a generic Australian Real Estate Investment Trust (A-REIT). How does your answer change if the A-REIT is a stapled security? Please include discussion of target leverage, maturity, debt security and the influence of common debt covenants. (500 words max). 3) [2.5 Marks] The ABC Corporation is considering opening an office in a new market area that would allow it to increase its annual sales by $2.85 million. The cost of goods sold is estimated to be 45 percent of sales, and corporate overhead would increase by $297,000, not including the cost of either acquiring or leasing office space. The corporation will have to invest $2.3 million in office furniture, office equipment, and other up-front costs associated with opening the new office before considering the costs of owning or leasing the office space. A small office building could be purchased for sole use by the corporation at a total price of $4.1 million, of which $700,000 of the purchase price would represent land value, and $3.4 million would represent building value. The cost of the building would be depreciated over 42 years. The corporation is in a 30 percent tax bracket. An investor is willing to purchase the same building and lease it to the corporation for $467,000 per year for a term of 15 years, with the corporation paying all real estate operating expenses (absolute net lease). Real estate operating expenses are estimated to be 48 percent of the lease payments. Estimates are that the property value will increase over the 15-year lease term for a sale price of $5.2 million at the end of the 15 years. If the property is purchased, it would be financed with an interest-only mortgage for $2,870,000 at an interest rate of 10 percent with a balloon payment due after 15 years. a. What is the return from opening the office building under the assumption that it is leased? b. What is the return from opening the office building under the assumption that it is owned? c. What is the return on the incremental cash flow from owning versus leasing? d. In general, what other factors might the firm consider before deciding whether to lease or own?
22LLP207: Research Methods SPSS 2: Descriptives and Internal Reliability 1. Download the file “P2_career_interests.sav” from the LEARN and open it using SPSS software. • The dataset contains responses to a “career interests” questionnaire completed by 250 people. People were asked to rate how interested they were in pursuing a career in this job role using a sliding ‘visual analogue’ rating scale running from very low (0) to very high (8). Data source: http://psych.colorado.edu/~carey/courses/psyc7291/ClassDataSets.htm • There is no missing data or reverse-coded items in this particular dataset • Also fefer to your lecture slides from Lecture 5 for detailed guidance on performing the SPSS analyses requested below. • Although we will not be entering and labelling data in this practical exercise, look carefully at how the dataset is set up. Data is entered simply by clicking on a cell and typing it in, just like Excel software, via the “Data View” tab (see the bottom left corner of the screen). Labels are entered by typing in the appropriate cells via the “Variable View” tab (see bottom left corner of the screen). 2. Calculate the means and standard deviations of each of the following variables and write them in the table below to two decimal places. Carpenter Forest Ranger Mortician Police Officer Fire Fighter Sales Representative Teacher Business Executive Stockbroker Artist Social Worker Truck Driver Doctor Vicar Actor Lawyer Architect Landscaper 3. Download the file “Graphs_LLP207.xls” from the LEARN and open it using Excel 4. Draw a line graph of the mean interest rating for each job role, using the appropriate table and graph template in the Excel file. (You may, if you wish, sort the job roles into numerical order of their mean interest rating before graphing them.) Label the graph and its axes. 5. Calculate the Cronbach’s alpha reliability coefficient for each of these psychometric scales and write these in the table below. 6. Compute an overall psychometric scale score for each job family from the mean of the constituent item ratings and save these as new variables, naming each with the appropriate job family label. 7. Calculate the mean and standard deviation for each psychometric scale score and write these in the table below also. · Lawyer · Mortician · Doctor · Stockbroker · Architect · Sales Representative · Business Executive · Landscaper · Forest ranger · Truck driver · Carpenter · Social worker · Teacher · Vicar · Police Officer · Firefighter · Actor · Artist 8. Draw a bar graph of the mean psychometric scale score for each job family, using the appropriate table and graph template in the Excel file. Label the graph and its axes.
School of Mathematics and Statistics MAST30021 Complex Analysis, Semester 1 2025 Written assignment 3 and Cover Sheet 1. Mandatory Summary 10 points. Write a summary of three lectures chosen from Lecture 16 to Lecture 20. Note, that any Theorem and Definition, especially those with a name of a renowned mathematician, are worth mentioning. Use the space below. Clearly indicate which lecture you are writing about. Lecture : 2. Question (simple computation) 10 points. Find all zeros and singularities of the following functions and classify those (isolated or not, essential, removable or the order of the poles and zeros). Give an explanation of your classification. Moreover compute the residues at all singularities where the residue is defined (removable singularities are ex-cluded). You are allowed to use the fact that all zeros of sin(z) in the complex plane lie at z = πn with n ∈ Z. Simplify your results as much as possible (fractions and factors of π remain as they are)!) (a) (you must make use of Landau symbols when computing the residue), (b) (you must make use of the limit formula when computing the residues). 3. Question (simple proof) 10 points. Consider the real function (a) Compute the Taylor series at any x0 ∈ R+ {1} and show that its radius of convergence is R(x0) = |1 − x0|. When computing the radius of convergence make use of the theorems and corollaries of Lectures 12 & 13 (say which you use and why you can apply those!) so that no ϵ − N criterion is required. (b) Prove that the union of the open discs of convergence D(x0, |1 − x0|) for x0 ∈ R+ {1} of these Taylor series is equal to the set 4. Question (advanced computation) 5 points. Compute the Laurent series at any point z0 ∈ C and any open annulus centred at z0 of the function Do not forget to state what the inner and outer radii of convergence Ri < Ro are? (You only need to give a simple explanation via holomorphicity of the function why these are the radii.) Hint: it is sometimes helpful to make use of the contour integral formula to find the Laurent coefficients! The total number of structurally different Laurent series you should find is 3. Moreover, you will need at some point the product rule for the nth derivative which is 5. Question (advanced proof) 5 points. Prove the following statement with the help of the theorems and statements up to Lecture 15: Let f(z) and g(z) be two entire functions satisfying |f(z)| ≤ |g(z)| for all z ∈ C, and there exists a z1 ∈ C such that f(z1) = g(z1) ≠ 0. Prove that f(z) = g(z) for all z ∈ C.
INFO1113 Assignment 1 - 2048 Due: 13 April 2025, 11:59PM AEST This assignment is worth 8% of your final grade. Task Description Implement the 2048 game using the Processing library for graphics and Gradle as a dependency manager. Setup of Gradle and a demonstration of the graphics capability of Processing will be done in the week 5 tutorial. You can access the documentation fromhere. As with any assignment, make sure that your work is your own, and do not share your code or solutions with other students. Figure 1. 2048 Example Game State Figure 2. 2048 Tile Colour Reference. Any value higher than 2048 uses the same colour as 2048. The game takes place on an n × n grid, initially with 2 randomly placed blocks (which are randomly selected from either a 2 or 4 with equal chance). The parameter n should be retrieved from command line arguments. If it is invalid or not provided, use the default which is n=4. Each turn, the player may make one of four moves with the arrow keys: UP, DOWN, LEFT or RIGHT. This shifts all blocks in that given direction, and also merges adjacent blocks with the same number into a higher value (eg, 4 and 4 makes 8). You should ensure that block movement is smoothly transitioning from one cell to the next (as in the example game linked above), OR that only one block moves one space per frame. After the pIayer’s move is finished, a new random bIock wiII spawn in one of the empty spaces. This can be either a 2 or a 4. In addition, the player may also spawn any amount of new blocks into empty cells themselves by clicking on an empty space, and a 2 or 4 will then spawn there (randomly chosen with 50% chance). If the player is unable to make any moves that change the position of blocks, the game is over. A timer in the top right corner of the screen should keep track of the number of seconds since the game began. When the game ends, this timer stops, and the text “GAME OVER” is displayed in the centre of the screen. The pIayer can press (r’ to restart the game. Marking Criteria (8%) Your final submission is due on Sunday 13 April 2025 at 11:59PM. To submit, you must upload your build.gradle file and src folder to Ed. Do NOT submit the build folder. Ensure src is in the root directory with the other files, and not part of a zip, then press MARK. Shown during tutorial in week 5 when Gradle and Processing are covered. (-1 deduction if your submission breaks these features) • • • Window launches and shows initial layout correctly with brown cells. Hovering over tiles with cursor changes the brown tile to a lighter colour to highlight it Left clicking on a cell pIaces a (2, or a (4, randomIy in that position 1 mark • • Board size adjusts based on command line argument input (default 4x4 if not provided) Timer counts up every second in the top-right corner of the screen. It stops when the game ends. 0.5 marks • Tile colours are correctly shown to differentiate different values 1.5 marks • Arrow keys result in tile movement in the corresponding direction (LEFT, RIGHT, UP, DOWN) if there is empty space for the tiles to move 0.5 mark deduction for each with bugs or not working 1.5 marks • Adjacent tiles are merged correctly after the player makes a move in either LEFT, RIGHT, UP or DOWN directions 0.5 deduction for each not working 1 mark • Random bIock spawns in an empty ceII after the pIayer,s move either a 2 or a 4 1 mark • If the player is unable to make any moves that change the position of blocks, the game is over. DispIay“GAME OVER”. PIayer can press (r, to restart the game. 1.5 mark • Movement is progressive such that blocks smoothly transition from one cell to the next (such as in https://play2048.co/) OR only one block can move one space on each frame.
Marketing 229 – Routes to Market Course Overview This module is a core element of Marketing degrees at Lancaster. It covers key asp[ects of modern marketing in considering point of sale (retail). It is directly relevant to what we shall term “Route to Market Marketing jobs” . These include shopper marketing, trade marketing and category management. The module enlarges your career options. Mktg builds on Mktg 101 or Mktg 227. It provides background relevant to final year modules including Strategy (301), Negotiations (303), Global Marketing (303) and Business to Business Marketing(329). It is a prerequisite for Marketing 327 and Mktg225. Module Aims Provide students with understanding of practical challenges in marketing posed by distribution. Introduce students’ knowledge and understanding of techniques in managing the routes through which consumer products reach audiences. Deepen understanding of value as understood by shoppers and organisations Enhance understanding of ethical issues in marketing. Deepen understanding of context in which marketers work. Module Learning Outcomes Understand the role of the route to market in marketing, how value is created and captured in this. Relate the availability of products to upstream activity and to the nature of the product. Identify channel functions and how these take place. Appreciate interconnections in the route to market – and their practical and ethical implications. Understand the importance of the shopper and the concept of shopper segmentation. Understand techniques used by brands in retail contexts. Employability Skills Conduct a process of enquiry- conduct independent research, evaluate sources, present information in a visually accessible way. Analyse and evaluate industry data. Concise communication of analysis and conclusions Autonomous work skills and ‘self management,. Module Structure The module comprises three themes with relevant information and discussion delivered in lectures. Theme 1: Understanding marketing at the retail level Week 11 – Introduction to the module and understanding VALUE in your own shopping Week 12 – Retail structures and introduction to shopper psychology and marketing responses. Week 13 – Route to Market marketing jobs – Shopper Marketing, Category Management and Trade Marketing. What do these jobs involve? Week 14 – by now your assignment should be quite advanced, so we will have aa lecture to look at information searches, use of information and AI anything else that allows you to relate your project to the lecture content. Theme 2 – Marketing and supply chains Week 15– Delivering value through the supply chain – how do we get what drives value in our personal value statement? What has to have happened behind the scenes? What do marketers need to understand about supple chain activity? Week 16 - Internationalisation and global supply chains. We will be looking at contemporary supply chains and there will be some politics discussed. We will be looking at tariffs or other issues of relevance at the time. Theme 3 – Ethics in the route to market. Week 17 – Global connections – we shall be looking at how you are connected to the world through the goods and services that you consume. What are the implications of this for social and environmental justice? Week 18 - How do marketers seek to use ethics to attract customers. To what extent do you trust the ethics of the products you buy? Week 19 – Bottom of the Pyramid Marketing and luxury markets We shall be looking at these contrasting markets and at the particularities of the supply chain and marketing activity in both cases. Week 20 – Review of the module and exam preparation. Workshops Workshops are a very important part of the module and take place in weeks 12, 14, 16 and 18. Brief preparation is required for each workshop. A file will be added in the preceding week to the workshop section of moodle. This includes brief reading and questions to prepare. We anticipate that this will take you about 1 hour. Workshop activity is the best place to test your understanding and to cover practical topics that are useful to the exam. Assessment – MKTG229 is assessed 50% by coursework and 50% by exam. Coursework In coursework you must pick any object that you have bought and that interests you (please no bananas!). Your project should explain how value has been delivered to you and according to your own ‘value statement,. The project is therefore personal. You will need to consider – why did you buy it and at that price? What was the influence of the retail environment on your purchase? How has shopper marketing or store placement been used to influence you? In order to deliver to your value statement (eg being fresh or being right place right time, or being cheaper than the alternatives) what activity has taken place behind the scenes and how was this managed? In other words, your task is to unpack the seemingly mundane and everyday task of getting goods into shoppers, hands. A lot of work goes into the mundane and it is your task to show that. Your assignment can be delivered in a variety of formats – but you are limited to 800 words (+/-10%). This word count is tight. But you are able to additionally use ANY visuals (eg graphs, cartoons, pictures, figures, timelines etc etc etc) in which you can convey information more effectively that through lengthy writing. Words included in your visuals do not count in your word count. Your assignment is due in at the end of term. More information will be given in a separate moodle document. This will explain also the criteria by which we assess your work. EXAM – there is also a take home summer exam. You will need to answer 2 of 4 questions. Fuller information will be issued in advance. Resources will be added to moodle – these may include academic papers, trade publications, videos, podcasts and more.
My experience with the Udemy-like online short course on Harnessing Artificial Intelligence for Penetration Testing has been equally a challenging and rewarding experience. One of the most significant hurdles I faced was overcoming my struggle to speak seamlessly during recording sessions. Even with meticulously prepared notes, I often stumbled over my words and lacked the fluidity I had originally envisioned for my course. I could spend hours studying a topic, but as soon as I began recording, my mind would go blank. This is something I have struggled with during interviews in the past, but to experience it when alone was one of the most internally frustrating situations I have dealt with.In an attempt to overcome this challenge, I decided to adopt a scripted approach. However, this brought forth its own set of challenges, as reading made my voice dull and not engaging. Despite numerous re-recordings, I acknowledge there’s still room for improvement, but I have started to find a better balance between being engaging and informative. This process pushed me mentally and emotionally, and I believe my communication and research skills have significantly improved over this semester.Although counterintuitive, I do hope for future opportunities to reattempt the creation of a short course. If I could do it over again, I would focus my research on quality over quantity, prioritize time management, and embrace imperfections on audio. Despite the challenges, I am grateful for this experience as I believe this semester has contributed significantly to my overall learning and personal development.
Instructions Based on the extended literature review and initial course recordings developed in previous assignments, complete the development of the Udemylike online short course on the approved topic. Record the remaining lectures, create supplementary materials, visual aids, and any additional resources. Ensure that the course is well-structured, engaging, and adheres to high educational standards. Include a reflection on the course development process, challenges faced, and potential future improvements. Assignment Notes Review Submission Requirements Ensure writing is academic and formal, avoid first-person perspective Submit your work in one single pdf file (Do not submit multiple files) There is no specific page limit, but literature review must be comprehensive and detailed Submission must follow IEEE Format Refer to IEEE Reference Materials for more information Utilize the Example IEEE Template Assignment Tips Homework 2-4: Consider finding recent and representative works related to your research. Look beyond research papers; explore technical blogs, white papers, and seminar/video talks. Explore notable venues in your research field (e.g., RSA, BlackHat conferences for security). Research involves not only stating your goals but also presenting your contribution clearly to a wide audience. Homework 2 or 3: Consider creating a visual representation (e.g., a figure or workflow) to help convey your proposed idea. Homework 2, 3, and 4: Pay attention to readability and generalizability. Ensure that you provide sufficient background information and detailed explanations to make your work approachable, understandable, and accessible Aim for your work to be accessible to both the general public and experts in your field.
Mid-Semester ExamSolution to Problem 1: Random-Maze Environment Implementation Problem 1: Notebook for problem 1 Problem 2: Notebook for Problem 2 Problem 3: Notebook for Problem 3 The given environment described below is a random maxe environment consisting of a grid of size (3 x 4), a highly stochastic environment with 11 states: 2 terminal states namely goal and hole states having rewards of +1 and -1 respectively. All the other non-terminal states has a reward of -0.04. For intended action, the agent has a 80% probability of taking that action and remaining 20% is distributed equally between every pair of orthogonal actions. The agent starts from state 8 with a a discoun factor of γ = 0.99. The figure(1) is shown below:Figure 1: Random Maze Environment 1. The evironment is created and the seed is set to 123. The trajectory generated based on the given environement constraints on setting the required seed is shown in the table below: Current State Action Reward Next State Done 8 2 -0.04 9 False 9 1 -0.04 10 False 10 1 -0.04 6 False 6 0 -0.04 6 False 6 0 -0.04 6 False 6 1 -0.04 2 False 2 3 1.0 3 True Table 1: Transitions table 2. The second trajectory generated by setting the seed to 8888 is shown below: State Action Reward Next State Done 8 1 -0.04 4 False 4 3 -0.04 8 False 8 0 -0.04 8 False 8 3 -0.04 9 False 9 3 -0.04 10 False 10 2 -0.04 11 False 11 2 1.0 7 True Table 2: Transitions Table Solution to Problem 2: RME Optimal Policy via Dynamic Programming The solution provides the optimal values for all the states calculated using Policy Iteration and Value Iteration algorithm. The major difference is in Policy iteration, the agent calculates the values for all states initially using an adversarial policy and then it improves the policy from the q function by choosing the action that maximizes the expected return in each state. These two processes are repeated until the values converge. Whereas for the case of Value Iteration, the agent calculates the maximum possible return for each action in each state and then dynamically updates the value of each state with that maximum value. 1. For this question, theta is taken to be 10e-10. The policy iteration firstly calculates the value functions for the given environment using the adversarial policy and then performs policy improvement by updating the Q values. The initial adversarial policy was taken randomly from the set of all possible actions. The adversarial policy used for the given Random Maze environment is shown below:Figure 2: Initial Adversarial Policy After performing policy iteration using the θ as 10e-10 and γ as 0.99, it was found that the optimal policy was reached within 3 iterations. The optimal policy for the Random Maze Experiment is illustrated below: The optimal state values for Policy Iteration are = [0.824, 0.892, 0.954, 0, 0.764, 0, 0.688, 0, 0.697, 0.639, 0.606, 0.381] 2. The second part is related to the implementation of Value iteration algorithm. This algorithm evaluates the value function until it converges to the optimal value function. Once the optimal values are obtained, the optimal policy for the environment can be calculated from them. After performing value iteration using the θ as 10e-10 and γ as 0.99, it was found that the optimal policy was reached within 25 iterations. The optimal policy for the Random Maze Experiment is illustrated below: The optimal state values for Value Iteration algorithm are = [0.824, 0.892, 0.954, 0, 0.764, 0, 0.688, 0, 0.697, 0.638, 0.606, 0.381]Figure 3: Final optimal Policy (Policy Iteration)Figure 4: Final optimal Policy (Value iteration) Solution to Problem 3: RME Prediction with MDP Unknown 1. The environment resets the agent to the state 8 of the all the nodes every time a new episode is initiated. The optimal state-value functions obtained though is independent of the starting point of the agent. The generateTrajectory() function generates a new episode containing the experience tuple (state, actions and indicators whether epsiode has terminated or not) and MC and TD plots are obtained y varying the step parameter implemented in the function decayAlpha(). 2. The plots for linear and exponential decay of the step parameter is shown below. The initial value is set to and final value is set to , max steps is set to 100(a) Linear Decay (b) Exponential Decay 3. The FVMC plots for all the non terminal states are shown below. From these plots, we can infer that the values for state 10 and 11 doesn’t change at all and the variations of all the state values are quite noisy especially because of the episodic nature of Monte Carlo algorithm. The peaks are comparatively lower in the case of EVMC as compared to FVMC algorithm and there is no significant difference between the values of the non terminal states in both the methodsFigure 6: FVMC plots for constant α = 0.5Figure 7: EVMC plots for constant α = 0.5Figure 8: TD plots for constant α = 0.5Figure 9: TD plots for constant α = 0.5 5. The next plots depicts the n Step TD algorithm where in which is the parameter that determines the number of steps that we want to look ahead before updating the value function. For the case of low n, the updates are based on the immediate reward and the estimated value of the next state, leading to high bias because it doesn’t consider the actual outcomes of future actions beyond the immediate next step also this leads to quicker convergence since for high values of n, the agent accumulates rewards over broader states and slowly high n step TD approaches the traditional MC algorithm.Figure 10: n step TD plots for constant α = 0.5 and n = 3 6. The next set of plots depicts the TD Λ algorithm. The plot depicts that the algorithm instantly converges to the optimal value which is indicated by no variations in the values of the states in the environment. 7. The next figures depicts the averaged plots for EVMC, FVMC, TD, 3 step TD and 6 step TD, TD (0.3) algorithms averaged over varying seeds ranging from 100 to 300.The averaged estimates are over500 episodes with α decaying from 0.5 to 0.01. These plot eliminates the noise and its clear that TD plots converge way more quickly with less variance than Monte Carlo estimates. 8. The next figures depicts the averaged plots for EVMC, FVMC, TD, 3 step TD and 6 step TD, TD (0.3) algorithms averaged over varying seeds ranging from 100 to 300.The averaged estimates are over500 episodes with α decaying from 0.5 to 0.01. These plot eliminates the noise and its clear that TD plots converge way more quickly with less variance than Monte Carlo estimates. The episodes are taken in the log scale which helps us to zoom into the initial stages of the learning process. 9. The FVMC-target and EVMC-target plots for state 4 is shown below. The plot depicts a higher variance in the beginning of episodes and the values finally becomes stationary which indicates that the learning is almost complete. The values don’t show an upward or downward trend which denotes that the algorithm is stable and not diverging and there are no sudden explorations in later episodes. Since, Monte Carlo estimation targets involves calculations of returns after an entire episode, the plot also denotes sufficient exploration of the environment by the agent. 10. The TD-target plot for state 4 is shown below. The target plot depicts that the the TD-targets initially have a high variance which denotes that the agent is beginning to explore and dynamics. The agent is updating its estimates over a wide variety of possible experiences. The faster convergence also denotes that the step parameter is comparatively good otherwise the values might have converged slower. Since, the target values come closer to each other, this indicates that the learning is almost complete and the targets have converged to a final valueFigure 11: n step TD plots for constant α = 0.5 and n = 6(a) FVMC Averaged out value estimates over 500 (b) EVMC Averaged out value estimates over 500 episodes episodes(b) 6 Step TD Averaged out value estimates over 500 (a) TD Averaged out value estimates over 500 episodes episodes 11. A comparative analysis of the algorithms has been incorporated into a bar plot depicting the optimal values for all of the non terminal states in the Random Mazke Environment. The figure is illustrated below:(a) 3 Step TD Averaged out value estimates over 500 (b) TD(0.3) Averaged out value estimates over 500 episodes episodes(a) FVMC Averaged out value estimates over 500 (b) EVMC Averaged out value estimates over 500 episodes (log scale) episodes (log scale)(a) TD Averaged out value estimates over 500 episodes (b) 6 Step TD Averaged out value estimates over 500 (log scale) episodes (log scale)(a) 3 Step TD Averaged out value estimates over 500 (b) TD(0.3) Averaged out value estimates over 500 episodes episodes(a) FVMC-Targets for state 4 (b) EVMC Targets for state 4Figure 19: TD Targets for state 4Figure 20: Optimal Values of states for different algorithms
