Defect Prediction Model for Object Oriented Software Based on Particle Swarm Optimized SVM

In terms of the security problem of power information system, this paper analysed the importance of the software defect prediction method in object-oriented software development, and proposed a software prediction model based on particle swarm optimized Support Vector Machine (SVM) corresponding to the features of object-oriented software. The model mainly consists of three parts: the first is the pre-processing module which normalizes the original data and selects feature, then the second is adaptive inertia weight particle swarm module which optimizes the parameters of SVM with the prediction accuracy as the fitness. Finally, the last SVM classification module predicts categories of reduced-dimension data using the optimal parameters from the second module. Experimental results show that the accuracy of the proposed model is 8.2%-12.2% higher than the comparative model, and 9.9%, 5.6% and 7.7% higher on the precision, recall and F value, which proves the validity of the proposed model.