Created by W.Langdon from gp-bibliography.bib Revision:1.8051
The DM procedure that uses a GP as a data mining function to create a subtree of UAVRM is discussed. The resulting decision logic for the RMs is rendered in the form of fuzzy decision trees. The fitness function, bloat control methods, data base, etc., for the tree to be evolved are described. Innovative bloat control methods using computer algebra based simplification are given. A subset of the fuzzy rules used by the GP to help accelerate convergence of the GP and improve the quality of the results is provided. Experimental methods of validating the evolved decision logic are discussed to support the effectiveness of the data mined results.",
Genetic Programming entries for James F Smith III ThanhVu Nguyen