CIS 490 - Machine Learning

3 credits
Prerequisites: CIS 360
Constructing computer programs that automatically improve with experience is the main task of machine learning. The key algorithms in the area are presented. Learning concepts as decision trees, artifical neural networks and Bayesian approach are discussed. the standard iterative dichotomizer (ID3) is presented, the issues of overfitting, attribute selection and handling missing data are disucssed. Neural nets are discussed in detail, examples of supervised and unsupervised learning are presented. Instance-based learning, i.e. k-nearest neighbor learning, case-based reasoning are introduced. Genetic algorithms are discussed on introductory level.