Code: BE3M33UI Artificial Intelligence
Lecturer: Ing. Petr Pošík Ph.D. Weekly load: 2P+2C Completion: A, EX
Department: 13133 Credits: 6 Semester: S
Description:
The course deepens and enriches knowledge of AI gained in the bachelor course Cybernetics and Artificial Intelligence. Students will get an overview of other methods used in AI, and will get a hands-on experience with some of them. They will master other required abilities to build intelligent agents. By applying new models, they will reiterate the basic principles of machine learning, techniques to evaluate models, and methods for overfitting prevention. They will learn about planning and scheduling tasks, and about methods used to solve them. Student will also get ackquainted with the basics of probabilistic graphical models, Bayesian networks and Markov models, and will learn their applications. Part of the course will introduce students to the area of again populat neural networks, with an emphasis to new methods for deep learning.
Contents:
1. The relation of artificial intelligence, pattern recognition, learning and robotics. Decision tasks, Empirical learning.
2. Linear methods for classification and regression.
3. Non-linear models. Feature space straightening. Overfitting.
4. Nearest neighbors. Kernel functions, SVM. Decision trees.
5. Bagging. Adaboost. Random forests.
6. Graphical models. Bayesian networks.
7. Markov statistical models. Markov chains.
8. Expectation-Maximization algorithm.
9. Planning. Planning problem representations. Planning methods.
10. Scheduling. Local search.
11. Neural networks. Basic models and methods, error backpropagation.
12. Special neural networks. Deep learning.
13. Constraint satisfaction problems.
14. Evolutionary algorithms..
Seminar contents:
Students will solve practical tasks. They will get experience with chosen packages for machine learning, graphical models, neural networks, etc. and will implement parts of algorithms themselves.
Recommended literature:
S. Russel, P. Norvig: Artificial Intelligence - A Modern Approach, 3rd ed., 2010
C.M. Bishop: Pattern Recognition and Machine Learning, 2006

Abbreviations used:

Semester:

Mode of completion of the course:

Weekly load (hours per week):