Function Sequence Genetic Programming for pattern classification

Pattern classification is one of the most researched problems in Artificial Intelligence. Genetic Programming (GP) has been used to construct classifiers by many researchers. Function Sequence Genetic Programming (FSGP) is a new variant of GP, base on which constructing classifier has not been investigated now. This paper explores the application of FSGP to pattern classification. Base on FSGP, binary classifier and multi-classifier are constructed. Experiments on four well-known data sets are made to demonstrate the classification performance of FSGP.