A Software Reliability Prediction Algorithm Based on MHPSO - BP Neural Network

Because the weights and thresholds of BP neural network usually adopt random assignment, there is a problem of low accuracy in software reliability prediction. In order to solve this problem, a software reliability prediction algorithm (MHPSO-BP) based on multi-layer heterogeneous PSO optimized BP neural network is proposed in this paper. In this algorithm, the population structure of the particle swarm is set to the hierarchical structure, and the velocity updating equation of the particle is improved by using the attractor. The information interaction between the particles is enhanced, and the optimization performance of the particle swarm optimization algorithm is improved. And then use the improved PSO to optimize the weight and threshold of the BP neural network. The software reliability prediction experiment was performed using the JM1 software defect data set of the NASAMDP project during the experiment. The results show that the proposed method has better predictive performance than the traditional BP neural network.