Learning Classifier System

A Learning Classifier System (LCS ) is a machine learning methods are combined with the evolutionary algorithms with classical learning algorithms to generate adaptive systems.

These systems are based on rules ( if-then ) have traditionally been the form of the conditional statement. These rules are to perform (input engl.) the best possible behavior due to a particular input. They are adapted with an evolutionary algorithm. In principle, in particular the Pittsburgh -type LCS where separate populations are adjusted by rules that differ from the Michigan Style LCS. In the latter, individual rules will evolve and not complete rule sets.