COURSE LEARNING GOALS The objective of the class is to:
(a) show how to identify the appropriate AI solutions for different classes of computational challenges and
(b) provide experience in implementing such solutions on representative challenges.
The course is intended for computer science graduate students, who have not been exposed to artificial intelligence material in the past. It can also appeal to students in related areas (such as psychology, mathematics, electrical, mechanical or biomedical engineering, etc.) who have interests in artificial intelligence methodologies and their applications.
INSTRUCTORAbdeslam Boularias
OFFICE HOURS Fridays 2:00-3:00 PM in CBIM 07
TEACHING ASSISTANTS
TOPICSThe class introduces fundamental ideas that have emerged over the past fifty years of AI research and provides a useful toolbox of AI algorithms. Example topics include:
(a) Deterministic Reasoning: Heuristic Search, Local Search, Adversarial Search, Constraint Satisfaction Problems
(b) Probabilistic Models: Bayesian Networks, Hidden Markov Models, Kalman and Particle Filters, (Partially Observable) Markov Decision Processes
(c) Machine Learning: Linear Models for Regression and Classification, Neural Networks, Kernel Methods, Gaussian Processes, Sparse Kernel Machines, Reinforcement Learning, Perception
BOOKS
Example textbooks include:
- "Artificial Intelligence: A Modern Approach" by Stuart Russell and Peter Norvig (Third Edition), Prentice Hall Series in Artificial Intelligence;
- "Pattern Recognition and Machine Learning" by Christopher Bishop, Springer
EXPECTED WORKRegular readings and homeworks, some of which involve programming, and exams.
EXAMS
A midterm and a final examination. Typically the midterm exam covers the material on deterministic reasoning, as well bayesian networks and inference. The final exam also covers material on Markov Decision Processes and machine learning.
GRADING SCHEME Midterm: 20% Final Exam: 20% Homework: 30% Final Project: 30%
TENTATIVE SCHEDULE (subject to changes) | 