|
Department of
Computational
Perception
|
|
SPEZIELLE KAPITEL AUS INTELLIGENTE SYSTEME:
MACHINE LEARNING
344.001, KV, 2std., WS 2005/06
Univ.-Prof. Dr. Gerhard Widmer
Institut für Computational Perception
Johannes Kepler Universität Linz
TIME: Tuesday, 12:00 - 13:30
START: Tuesday, 11.10.2005
PLACE: HS 13 (TNF Tower, Ground Floor)
NOTE: This class will be taught in English!
Goals:
Machine Learning is a sub-field of Artificial Intelligence that deals with the
development of computer programs that can learn in some sense. Learning is a
very broad and multi-faceted concept that comprises all kinds of processes that
derive general knowledge from observations and experiences. Machine Learning
is an extremely active field of basic research, but many learning algorithms
are also playing an increasingly important role in practical applications of
intelligent systems - e.g., learning robots, adaptive user interfaces,
automatic recognition and classification systems, etc. Another related and
very important field of application of learning algorithms is Data Mining -
the discovery of hidden and useful patterns and relationships in huge amounts
of data.
The class gives an introductory overview of the basic concepts and most
important methods of Machine Learning. After one semester, the students should
have a basic understanding of the properties and limitations of machine
learning algorithms, and should be able to critically judge and evaluate
new research results in this field.
Contents:
The following topics will be covered in class:
- Basic notions in concept learning: generalisation, version space, bias
- Algorithms for learning classification rules and decision trees from
examples
- Instance-based and case-based learning
- Relational Learning and Inductive Logic Programming
- Autonomous generation of categories and taxonomies (clustering)
- Reinforcement learning
- Basics of evolutionary learning algorithms
- Basics of computational learning theory
- Application possibilities for machine learning
Course Materials:
pdf versions of the powerpoint slides used in class will be made available
via the Web (KUSSS).
Recommended book (but not needed to pass):
Mitchell, T.M. (1997). Machine Learning.
New York, N.Y.: McGraw-Hill.
Questions, Suggestions, Complaints, etc. to:
Gerhard Widmer
Tel. 2468-1510
gerhard dot widmer at jku dot at
last
edited by gw on
2005-09-01