Software defect prediction method based on fuzzy integral

Software defect can cause serious consequence, which results in the research hotspot about software defect prediction. The paper studys and discusses the application of fuzzy measure and fuzzy integral in the software defects classification, selects several commonly used classifiers to learn the software defect data, and sets up multiple classifier fusion model based on the theory of the fuzzy integral. Experimental results show that there are interactions between classifiers, and prove the multiple classifiers fusion technology based on fuzzy integral. Compared to single classifier, the proposed defect prediction method based on Fuzzy Integral can improve software defect prediction accuracy rate.