The use of cultural algorithms with evolutionary programming toguide decision tree induction in large databases

摘要

In this paper, we use an evolutionary computational approach basedupon cultural algorithms to guide the incremental learning decisiontrees by ITI. The results are compared to those produced by ITI itselffor a complex real-world database. The results suggest that ITI canindeed produce optimal trees in some cases, and can produce optimaltrees using an evolutionary approach in others