Abstract:

Gene Expression Programming (GEP) is an automatic programming approach widely used in many areas,such as time series analysis,classification,multiobjective optimization and massive data analysis.A new GEP algorithm is proposed by combining the advantages of the depthfirst and breadth-first technologies in the GEP decoding process.The new algorithm can increase the diversity of individuals and properly preserve better sub_ETs.The experimental results show that,compared with the standard GEP algorithm,the new algorithm can improve the mean fitness without increasing too much evolutionary time,thus achieving a higher success rate.

Key words: depth;breadth;decoding;GEP;symbolic regression