LAUSR

Data classification is an important task in the field of data mining, which can be used to mine the model of important data and forecast the future trend of those data. Although some breakthroughs have been made in data classification theoretically and technically, there are still some problems, such as lack accuracy of classification modeling algorithm, poor comprehensibility of classification rules and so on. Accuracy improvement and accurate achievement of classification has become hot research topics. Gene expression programming (GEP) has been considered a powerful evolutionary method for data classification. Aiming at the shortage of basic GEP classification algorithm, a novel classification algorithm based on GEP named O_GEPCA has been proposed in this paper. By using this method the initialization and mutation operator adjustment method, calibration set, evolution function and correction strategy will be improved, and the basic GEP classification algorithm will be optimized. The proposed O_GEPCA method shows significantly improvement after comparative study between our proposed O_GEPCA methods and the primitive GEP. The efficiency and capability of our proposed O_GEPCA for data classification will be tested in four well-studied benchmark test cases including card, cancer, heart, glass classification problem demonstrate.