Classification Based on Multi Expression Programming
Jinglei Guo · Journal of Chinese Computer Systems · 2010
Multi Expression Programming(MEP) is a linear structured genetic programming with reused genes,which has been applied in many data mining problems.However,less research work is focused on the classification problems based on MEP.A new classification approach is proposed combining MEP and other classification methods.The classification rules are embedded in the chromosome and evolved by survival of the fittest.The experiment indicates that the new approach is feasible and holds high classification accuracy.