Modified Himmelblau Function Classification with rGCS System
Łukasz Cielecki, Olgierd Unold · 2008
Learning Classifier Systems (LCSs) have gained increasing interest in the genetic and evolutionary computation literature. Many real-world problems are not conveniently expressed using the ternary representation typically used by LCSs and for such problems an interval-based representation is preferable. The new model of LCS – so-called rGCS – is used to classify real-valued data. In order to handle effectively with complex 3D functions, rGCS was extended by the covering technique and the co-populations in environment probing rules.