The content of the lecture is an introduction to the issues of artificial
intelligence, such as knowledge representation, heuristics, inference, action
planning, decision making, behavior under uncertainty, as well as an
introduction to the methods of machine learning.
Program:
1. State space representation, search, greedy strategies, the use of heuristic information, searching graphs, search for two-person games, and for the CSP problems.
2. Logic-based representation, theorem proving and reasoning using first order logic. The situation calculus. Reasoning under uncertainty, non-monotonic logic, truth maintenance systems (TMS).
3. Classical action planning: STRIPS and PDDL representations, POP and GRAPHPLAN algorithms.
4. Probabilistic representation: conditional probability, probabilistic belief networks, utility functions, the value of information.
5. Making complex decisions, Markov decision problems, basic algorithms, POMDP problems.
6. Machine learning methods: inductive learning, decision trees. Reinforcement learning. Computational learning theory, the PAC model.
Literature
* Russell, Norvig, Artificial Intelligence A Modern Approach, Third Edition, Prentice-Hall, 2010
* T.Mitchell, Machine learning, McGraw Hill, 1997
* Internet resources
