作者: Nada N. Saleem , Rita G. Al gargoor
DOI:
关键词:
摘要: Due to the growth in demand for software with high reliability and safety, prediction becomes more essential. Software is a key part of quality. Over years, many models have been successfully utilized practical engineering, however, no single model can obtain accurate all cases. So order improve accuracy proposed combine neural networks (NN). Particle swarm optimization (PSO) algorithm has chosen applied learning process select best architecture network. The applicability demonstrated through three failure data sets. results show that good capability applicable prediction.