Abstract: Aiming at the problem that the software system is becoming more and more complex and the software defect is difficult to detect,a software defect prediction model based on artificial immune recognition system was proposed.The model was constructed firstly by using the principal components analysis method to reduce the dimension of the original data set,and then the affinity between antibody and antigen was calculated based on the Gauss radial basis function (RBF).Based on affinity calculation,antibody training,resource competition and the selection of memory cells were conducted.Classification was performed using memory cell set at last.Simulation shows that our model prediction accuracy reaches 84%~90%,accuracy reaches 85%~91%.

Key words: Software defect detection,Artificial immune recognition system,Principal components analysis,Static analysis,Supervised machine learning