Abstract
Extracting accurate and understandable classification rules from data is a fundamental data mining activity. Fuzzy classification rules is considered a better classification of knowledge that the fuzzy rules of readability and analytical, and the use of fuzzy rules is very intuitive. In this paper, we present an improved gene expression programming (GEP) for extracting fuzzy classification rules by a logical operators instead of mathematical ones to represent the chromosome validity evaluation. Moreover, a novel technique to evaluate the fitness of the individual rather than transform the chromosome into expression tree is proposed. Our proposed approach has been tested on some benchmark datasets selected from UCI and the results indicate that our approach is highly comparable with other techniques including basic GEP, C4.5 and C4.5 rule.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call