The Research On Function Mining Based On Dynamic Evaluate Strategy Of GEP Algorithm

The information society made people attach more and more attention to data, so people are eager to discover the rules implied in the data. Therefore, proposing a high efficiency and exact way to execute function mining is a research emphasis in data mining. Evolutionary computation has highly developed in function mining and Gene Expression Programming (GEP) is a global optimization search technology developed in the past few years, and has been applied in many areas because of its powerful search capabilities and high evolution efficiency.This paper reviewed research status of GEP and evolutionary computation; summarized development, basic characteristic and branch of evolutionary computation; introduced each essential technique of GEP which mainly included the construction of genes and chromosomes in GEP and the design of fitness function and genetic operations; elaborated gene expression programming characteristics, which is better than GP and GA.In order to improve the efficiency of the GEP discovering knowledge, this paper put forward an improved GEP algorithm, proposed a technology of function mining, which based on dynamic evaluate method. Applied the stepwise adaptation of weights to the dynamic adaptation of GEP fitness function, with which improved the computing efficiency. This paper proposed the algorithm of the stepwise adaptation of weights. In order to verify the accuracy and the validity about GEP-SAW algorithm, this paper applied GEP-SAW to function mining, implemented the software based on dynamic evaluate strategies of GEP-SAW algorithm. Through mining one variable function, two or three variables function and the standard data, those experiment results indicated GEP-SAW has good performance and is superior to traditional algorithm and classical GEP, the models set up by GEP-SAW are better than the models set up by traditional algorithm and classical GEP, and its prediction accuracy is high.