代做Syllabus for Fa24 - PLAN/SRCH/RSN UNDR UNCERTY-WB (49964)代做Python编程

日期：2024-09-15 06:40

Syllabus for Fa24 - PLAN/SRCH/RSN UNDR UNCERTY-WB (49964)

Course Syllabus

Overview

This course will serve as an introduction to three key foundational problems inAI: planning, search, and reasoning under uncertainty. We will investigate how to define planning domains, including representations for world states and actions, covering both symbolic and path planning. We will study algorithms to efficiently find valid plans with or without optimality, and partially ordered, or fully specified solutions. In transitioning from classical to modern approaches to planning, we will cover decision-making processes and their applications to real-world problems with complex autonomous systems. We will investigate how in planning domains with finite state lengths, solutions can be found efficiently via search. Finally, to effectively plan and act in the real world, we will study how to reason about sensing, actuation, and model uncertainty. Throughout the course, we will relate how classical approaches provided early solutions to these problems, and how modern machine learning builds on, and complements such classical approaches.

Prerequisites

· Python programming

· Basic data structures and algorithms

· Introductory statistics

Optional Textbooks

The course materials are designed to be self-sufficient, but here are some optional textbooks that cover the material in the course:

· Artificial Intelligence: A Modern Approach, by Stuart Russell and Peter Norvig. https://aima.cs.berkeley.edu/ (https://aima.cs.berkeley.edu/)

· Probabilistic Robotics, by Sebastian Thrun, Wolfram Burgard, and Dieter Fox. http://www.probabilistic-robotics.org/ (http://www.probabilistic-robotics.org/)

· Planning Algorithms, by Steven M. LaValle. http://lavalle.pl/planning/ (http://lavalle.pl/planning/)

Course Materials

The course materials include:

· Lecture videos, available on Canvas

· Lecture slides, linked in the outline below

· Example questions, included every week on Canvas

· Assignment handouts, included as links on Canvas

Course materials are not to be shared outside the course without explicit instructor approval.

Breadth/Depth Materials

This course introduces a lot of material, and to aid in studying for the quizzes and assignments, each lecture has been designated as  a "depth" or "breadth" topic (see course outline below). Depth topics cover key algorithms central to the course, and you are expected to understand them in detail. The level of understanding must be such that you should be confident in being able to implement it

precisely - in fact, the programming assignments are drawn from these topics. The quizzes will include questions on algorithmic steps of depth topics, so it is important to understand them in detail. The breadth topics, on the other hand, provide a broader perspective,   and include a number of extensions to the basic algorithms - you will be expected to understand them well enough to answer

questions about general algorithmic properties (e.g., is this planning algorithm complete, optimal, or probabilistically complete?), but we will not test at the level of detail of being able to implement them.

https://utexas.instructure.com/courses/1404599/assignments/syllabus

Course Timeline

This course is designed to be self-paced, with regular checkpoints. The quizzes and assignments provide hard checkpoints and deadlines by which you must cover the relevant materials and submit the requisite evaluation materials. However, you are free to read ahead and work ahead of the schedule - we in fact encourage you to do so if you feel you have already mastered previous material well in advance of the due dates.

Grading

You will be graded on the basis of four quizzes, and four programming assignments. Each quiz and programming assignment will be worth 12.5% of the total grade. For quizzes, There will be five quizzes. the one with lowest grade in first three quizzes will be dropped. Quiz4 and Quiz5 will always be counted to final grade.

The final letter grading range: A: 94-100, A-: 90-93, B+: 87-89, B: 84-86, B-: 80-83, C+: 77-79, C: 74-76, C-: 70-73, D+: 67-69, D; 64- 66, D-: 60-63, F: 0-59. There is NO curve on final grading.

Programming Assignments

There will be four programming assignments. The objective of the programming assignments is to demonstrate the ability to implement key algorithms covered in the course. All programming assignments are to be done individually - that is, all code submitted must be written by you on your own. However, you are welcome to discuss the strategies and algorithms behind how to implement them with your fellow classmates. When discussing programming assignments, you are welcome to meet in groups and discuss ideas on (physical or virtual) whiteboards, as long as no code is shared.

Quizzes

There will be four quizzes and one make-up quiz in the course. Quizzes will be host in Canvas. Each quiz is timed and will be auto closed in 60 minutes after it starts. The objective of the quizzes is to demonstrate an understanding of the concepts and theory behind the topics covered in the course. These are closed-book, closed-notes quizzes, to be completed individually, and with no consultation  with any reference - person, electronic, written, or otherwise. The practice questions provided for every week are examples of types of questions that will be asked in the quizzes.