6.825: Techniques in Artificial Intelligence--Fall '02

Announcements Homework 3 Solutions (pdf) have been posted. The midterm solutions have been posted. Graded projects and midterms (except 2a) are available during office hours. A histogram of the midterm scores is available. Three printed chapters on probability and Bayesian networks are now available outside of NE42-759. Nilsson chapter 20 and chapters 14 and 20 of AIMA 2nd edition. References Textbook: Artificial Intelligence: A Modern Approach by Russell and Norvig. (Buy it at the Coop or online.) Broad background resource: AI Topics Java Resources Syllabus Includes pointers to required reading not in the textbook and suggested exercises. Description 6.825 is a graduate-level introduction to artificial intelligence. Topics include: representation and inference in first-order logic; modern deterministic and decision-theoretic planning techniques; basic supervised learning methods; and Bayesian network inference and learning. Prerequisites 6.042 (Mathematics for Computer Science) 6.046 (Introduction to Algorithms) (desirable, but not required) 6.034 (Artificial Intelligence) (desirable, but not required) Students should be familiar with uninformed search algorithms (depth-first and breadth-first methods), discrete probability (random variables, expectation, simple counting), propositional logic (boolean algebra), basic algorithms and data structures, basic computational complexity, and basic calculus. Students should also be aware that course assignments will require the use of the Java programming language. Grading The work for this course will consist of 4 take-home project assignments and two exams. The projects will count for 50% of the grade, and the exams, 50%. Late Policy for Projects: 10% off for each calendar day late. No credit if more than 5 days late. Collaboration We want to strongly encourage collaboration as a way for students to come to understand the material better. You may do the projects in groups of two, turning in a single write-up (or you may do it on your own). However, you may not partner with the same person for more than three project assignments. Projects 1a, 1b, and 1c count as distinct assignments. If you are looking for a partner for an assignment, email the class list asking if anyone is available. You are also quite welcome to discuss the assignments as much as you'd like between groups. The ultimate requirement is this: Don't put your name on anything you don't understand. There will, of course, be no collaboration allowed on the exams.