Aim of the course
The primary objective is to present the concepts, methods and algorithms related to the representation and processing of knowledge and the generation of knowledge from data.
Acquisition of knowledge from experts and associated problems. Knowledge representation using symbolic methods (classical logic, modal and temporal logic, fuzzy logic, description logic, ontologies, logic of trustable reasoning.) Processing methods of knowledge (automatic inference algorithms, query processing). Knowledge engineering languages: Lisp (introduction to the language and usage to build knowledge-based systems), Prolog (logical basics, concept of language). Selected problems in machine learning: basic concepts, learning under supervision of (learning of concepts, induction rules, generation of probabilistic models), learning with reinforcement (Markov decision processes, Q-learning, SARSA, extension of algorithms), inductive logic programming, transforming a set of attributes (selection , discretization). Examples of practical applications of these techniques.
Overview of the course elements
The course involves laboratory classes, which aim at a practical illustration of the issues addressed at the lecture. During the lab students will design and implement programs relating to the representation and processing of knowledge. Also they learn the Weka machine learning system and a selected library for the learning with reinforcement.
1. S. Russell, P. Norvig, Artificial Intelligence: A Modern Approach (3rd Edition), Prentice Hall, 2010
2. P. Cichosz: Systemy uczące się, Wydawnictwa Naukowo-Techniczne, 2000
3. J. J.Mulawka: Systemy ekspertowe, Wydawnictwa Naukowo-Techniczne, 1996
4. L. Rutkowski: Metody i techniki sztucznej inteligencji, Wydawnictwo Naukowe PWN, 2005
5. Handbook of Logic in Artificial Intelligence and Logic Programming, D.M. Gabbay, C. J. Hogger, J. A. Robinson, Oxford University Press, 1993-1994