Software Fault Prediction Using Artificial Immune Systems

Software fault prediction based on mining of code and design metrics has been considered by many researchers. Fault detection systems predict faults by using software metrics and data mining techniques. Various classifiers have already been used in this case; from mathematical to evolutionary algorithms. In this paper, we will present a new evolutionary approach for predicting software faults. This method uses negative and positive selection algorithms which are evolutionary algorithms inspired by natural immune systems to create some prototype for each class of defective and non-defective to determine whether an input data is defective or not? Then we will test it on NASA software fault dataset, review the results and show its advantages to other methods.