辅导案例-COMP90054 AI

  • May 15, 2020

The University of MelbourneSchool of Computing and Information SystemsCOMP90054 AI Planning for AutonomyProject 1, 2019Deadline: Monday 2 September 18:00This project counts towards 10% of the marks for this subject.This project must be done individually.AimsThe aims of this project are to improve your understanding of the various search algorithmsand to experience how to derive heuristics, using the Berkely Pac Man framework.https://inst.eecs.berkeley.edu/~cs188/fa18/project1.htmlYour taskYour tasks relate to the assignment at https://inst.eecs.berkeley.edu/~cs188/fa18/project1.html.Set Repository to Private (0 marks)Once you have forked the repository, your repository may be viewable by other students inthe class. To avoid any issues with academic misconduct, please set your repository to ’pri-vate’.You can does this by going to gitlab.eng.unimelb.edu.au, selecting your comp90054-a1-2019repository, navigating to the privacy settings using ‘Settings’ , then ‘General’, then ‘Permis-sions’, and selecting ‘private’.Please do this as soon as you fork the repository.Practice Task (0 marks)To familiarise yourself with basic search algorithms and the Pacman environment, it is agood start to implement the tasks at https://inst.eecs.berkeley.edu/~cs188/fa18/project1.html, especially the first four tasks; however, there is no requirement to do so.1Part 1 (2 marks)Implement the Iterative Deepening Search algorithm discussed in lectures. You should beable to test the algorithm using the following command:python pacman.py -l mediumMaze -p SearchAgent -a fn=idsOther layouts are available in the layouts directory, and you can easily create you own!Part 2 (2 marks)Implement the Weighted A* algorithm discussed in lectures using W = 2. You may hardcodethis weight into your algorithm (that is, do not pass as a parameter).You should be able to call your function using the fn=wastar parameter from the com-mand line, i.e. you should be able to test the algorithm using the following command:python pacman.py -l bigMaze -z .5 -p SearchAgent-a fn=wastar,heuristic=manhattanHeuristicOther layouts are available in the layouts directory, and you can easily create you own!Part 3 (6 marks)Now we wish to solve a more complicated problem. Just like in Q7 of the Berkerley PacMan framework, we woud like to create an agent that will eat all of the dots in a maze.Before doing so, however, the agent must eat a Capsule that is present in the maze. Yourcode should ensure that no food is eaten before the Capsule. You can assume that there isalways exactly one Capsule in the maze, and that there will always be at least one path fromPacman’s starting point to the capsule that doesn’t pass through any food.In order to implement this, you should create a new problem called CapsuleSearchProblemand a new agent called CapsuleSearchAgent. You will also need to implement a suitablefoodHeuristic. You may choose to implement other helper classes/functions. You shouldbe able to test your program by running the following code:python pacman.py -l capsuleSearch -p CapsuleSearchAgent-a fn=wastar,prob=CapsuleSearchProblem,heuristic=foodHeuristicAn agent that eats the capsule then proceeds to eat all of the food on the maze will receive 3marks. The remaining 3 marks will be based on the performance of your agent (i.e. numberof nodes expanded), as in Q7 of the Berkeley problem. Since you are using the Weighted A*algorithm, however, the number of node expansions required for each grade will vary.HINT: Think carefully about how you intend to structure your solution before startingto implement it.NOTE: You should not change any files other than search.py and searchAgents.py. Youshould not import any additional libraries into your code. This risks being incompatible withour marking scripts.2Checking your submissionRun the command:python autograder.pyto run the tests from the Berkely assignment, but also two additional sanity checks for ourtests, called comp90054-part1 and comp90054-part-2. These are very simple tests of yourweighted A-star and iterative deepening algorithms that should pass, but it is important thatyou are able to run the autograder and have these tests pass, otherwise, our marking scriptswill NOT work on your submission.For part 3, this will be marked using the command-line commands listed in the abovetasks.Marking criteriaThis assignment is worth 10% of your overall grade for this subject. Marks are allocatedaccording to the breakdown listed above, based on how many of our tests the algorithmspass. No marks will be given for code formatting, etc.SubmissionThe master branch on your repository will be cloned at the due date and time.From this repository, we will copy only the files: search.py and searchAgents.py. Donot change any other file as part of your solution, or it will not run. Breaking these instruc-tions breaks our marking scripts, delays marks being returned, and more importantly, givesus a headache.Note: Submissions that fail to follow the above will be penalised.Academic MisconductThe University misconduct policy1 applies. Students are encouraged to discuss the assign-ment topics, but all submitted work must represent the individuals understanding of thetopic. The subject staff take academic misconduct seriously. In the past, we have prosecutedseveral students that have breached the university policy. Often this results in receiving 0marks for the assessment, and in some cases, has resulted in failure of the subject.Important: As part of marking, we run all submissions via a code similarity comparisontool. These tools are quite sophisticated and are not easily fooled by attempts to make codelook different. In short, if you copy code from classmates or from online sources, you riskfacing academic misconduct charges.But more importantly, the point of this assignment is to have you work through a series offoundational search algorithms. Successfully completing this assignment will make the rest1See https://academichonesty.unimelb.edu.au/policy.html3of the subject, including other assessment, much smoother for you. If you cannot work outsolutions for this assignment, submitting another persons code will not help in the long run.4

LATEST POSTS
MOST POPULAR

ezAce多年来为广大留学生提供定制写作、留学文书定制、语法润色以及网课代修等服务,超过200位指导老师为您提供24小时不间断地服务。