CS 440: Introduction to Artificial Intelligence, 2025 Fall

Course overview

• We will cover the following topics in Artificial Intelligence
  • search
  • knowledge representation and reasoning
  • planning
  • reasoning under uncertainty
  • machine learning
• Instructor: Lirong Xia <lirong.xia@rutgers.edu>. Please call me Lirong and contact me using your Rutgers email.
• TAs: Doga Diren<dd1269@scarletmail.rutgers.edu>; Yunran Yang<yunran.yang@rutgers.edu>
• PT : Pradhyumna Kiledar <pk811@scarletmail.rutgers.edu>
• Time and location: Monday and Wednesday, 7:30pm-8:50pm, ARC 103
  • Lecture notes and slides are just for your reference. Students are responsible for coming to class and taking their own notes. Please login OneDrive with your Rutgers NetID and password, and do not share the materials outside this class.
• Office hours (see Piazza)
  • For all non-confidiental questions please use Piazza first. You will get timely help and everyone will benefit from your questions.

Tentative schedule (subject to change)

Week Date Topic Book chapter Other information 1 9-1       9-3 Introduction to the course Chapter 1, 2 Link to lecture notes and slides 2 9-8 Project 0 due by midnight Uninformed Search (BFS, DFS, Greedy) 3.4-3.4.3; 3.4.5   9-10 Informed Search (A*) 3.4.4; 3.4.6; 3.4.7; 3.5-3.5.2; 3.6   3 9-15 Informed Search (A*) 3.4.4; 3.4.6; 3.4.7; 3.5-3.5.2; 3.6   9-17 Alpha-beta pruning, Expectimax search 5   4 9-22 Project 1 due by midnight Alpha-beta pruning, Expectimax search 5   9-24 Constraint Satisfaction Problems 6   5 9-29 Project 2 due by midnight Probability, conditional independence 14.1-14.3   10-1 Exam 1 Review     6 10-6 In-class Exam 1     10-8 Bayesian network 1: definition, conditional independence 14.1-14.3   7 10-13 Bayesian network 2: inference, variable elimination 14.4-14.5   10-15 Utility 16.1-16.3   8 10-20 Written HW1 due by midnight Markov Decision Processes (MDPs) 21   10-22 Reinforcement learning 1   9 10-27 Reinforcement learning 15.2, 15.5   10-29 Written HW2 due by midnight Probabilistic Reasoning over Time 15.2, 15.5   10 11-3 Hidden Markov Models: Filtering Algorithm 15.2, 15.5, 15.6   11-5 Project 3 due by midnight Exam 2 Review     11 11-10 In-class Exam 2     11-12 Hidden Markov Models: Particle Filters     12 11-17 Speech Hidden Markov Models: Viterbi Algorithm     11-19 Project 4 due by midnight Naive Bayes     13 11-24 Perceptrons     11-26 no class (Friday schedule)       14 12-1 MIRA, SVM, and k-NN     12-3 Neural Networks     15 12-8 Project 5 due by midnight Advanced Topics     12-10 Final Exam (comprehensive) Review     16 12-15 In-class Exam 3 starting at 08:00PM

 

Textbook for reference (not required)

Artificial Intelligence: A Modern Approach (third edition), Prentice Hall, 2009. By Stuart Russell and Peter Norvig

Prerequisites

• A qualified student must have taken courses on
  1. Linear algebra
  2. Algorithm design
  3. Probability or statistics
• We will have programming assignments using Python, so you should be comfortable with using it

Objectives

• We hope to help with developing
  1. A taste of the field of AI
  2. Knowledge and techniques in major topics in AI
  3. Programming skills for implementing AI

General Class Policies

• Website and Announcements. We will make extensive use of piazza and the course website. You are responsible for checking the course website regularly for announcements and course materials, as well as your e-mail for communications related to the class.
• Lectures. Students are highly encouraged to attend all classes and are responsible for all material covered and announcements made in lecture.
• Laptops and Electronic Devices. No laptops or other electronic devices are allowed in lecture, excpet that you use it for in-class polling. Even if you are diligently taking notes on your laptop, the bright screen and the activity is extremely distracting to the people behind you. Because of this, "movie theater" rules apply: no laptops, phones, or other devices with a screen on them should be out during lecture, except during possible in-class polling. After the polling, you must turn it off. Students who continually disrupt the class will be asked to leave.

Grading

• Exam 1: 15%
• Exam 2: 15%.
• Final Exam (comprehensive): 30%.
• Projects: 30%
• Written Homeworks: 10%

Regrading Policies

• All grades are finalized one week after the release dates
• To submit a regrading request, please email Lirong and both TAs a formal proof of the correctness of submitted HW, project, or exam. Please prepare the proof by yourself, not by GenAI. We will read the proof and regrade the question accordingly.
• Please come to TAs' office hours for solutions of HW, projects, or exams. We will not distribute the solutions.

Project assignments and written homeworks

• There will be 6 project assignments. You can work in pairs if you want, but you must do the projects and submit them by yourself. Using ChatGPT or similar tools are strictly forbidden. Intro to AI is not a course on programming nor on algorithms. If you need help on projects, please ask on Piazza or contact TAs and PTL. Lirong does not help with Python or debugging projects, unfortunately.
• Evaluation: Your code will be autograded for technical correctness. Please do not change the names of any provided functions or classes within the code. However, the correctness of your implementation -- not the autograder's judgements -- will be the final judge of your score. If necessary, we will review and grade assignments individually to ensure that you receive due credit for your work.
• We also have some written homeworks. Unlike projects, you must do all written HWs by yourself, not in pairs or in groups. Discussions are allowed but you must describe whom you have discussed with and how you benefited from the discussions.
• Please use the methods taught in the class in all projects, HWs, and exams. Using other methods will result in 0 for that question. Innovations are very welcome and encouraged. Students are encouraged to contact Lirong via email or come to office hours to discuss novel methods and ideas.
• The bottom line is: Projects and HWs are the best practices for exams.

Academic dishonesty and late policy

• Student-teacher relationships are based on trust. For example, students must trust that teachers have made appropriate decisions about the structure and content of the courses they teach, and teachers must trust that the assignments that students turn in are their own. Acts which violate this trust undermine the educational process. In this class, you are allowed (and encouraged) to discuss homework and projects with other members of the class, and to formulate ideas together. However, everyone must write up their assignments completely separately, and include the names of everyone you discussed the assignment with. You may not copy (or near-copy) a solution from another, or use resources other than the class notes or the class textbook. Use of materials other than the class textbooks, including any material found on the Internet or material from previous versions of this course, is a clear breach of academic integrity and will be punished severely. No collaboration, or any electronic devices, is allowed during exams. Violating the above policy will result in the final grade being reduced by a letter and a 0-grade for the assignment for both parties. Depending on the circumstances, harsher penalties may be used, including a failing grade for the class.
• Due time. Project assignments and written exams must be turned in by midnight (11:59 pm) of the due date.
• Late policy. For each project or HW, if the submission is late but no more than 24 hours, then 25% penalty will be applied; if the submission is late between 24 hours and 48 hours, then 50% penalty will be applied. After 48 hours, no submission will be accepted.
• Regrading. Problems about grading must be reported within one week after the grades are made available. This applies to all homeworks, projects, and exams. After that, all grades are finalized and will not change unless the student can present an official excuse letter to explain why the request was not submitted in time.

Acknowledgements

Thanks Pieter Abbeel, Vincent Conitzer, John DeNero, Dan Klein, Malik Magdon-Ismai, Xintong Wang, Peter Sone for offering tremendous helps on developing the course!