Fuzzy decision tree construction with gene expression programming

Gene expression programming(GEP) is a kind of phenotype/genotype based evolutionary computation technique. GEP has been used to evolve parsimonious decision tree with high accuracy comparable to C4.5. However, basic GEP based decision tree can still be improved with fuzzy information processing. In this paper, we proposed a GEP based decision tree technique with fuzzy attributes transformation strategy. The technique can significantly reduce the number of branches need to split the decision tree, thus improve the classification accuracy. Results show that our method can find better generalized ability rules.