Comprehensive Evaluation of Software Quality Based on LM-BP Neural Network
Anbang Wang, Lihong Guo, Yuan Chen, Junjie Wang, Yuanzhang Song · 2017
In view of the shortcomings of traditional software quality evaluation methods, such as subjective and lack of self-learning ability, a comprehensive evaluation method of software quality based on Levenberg Marquardt Back Propagation (LM-BP) neural network is proposed. Comprehensive evaluation system of software quality is established based on ISO/IEC 9126 software quality model, and LM-BP algorithm is used to solve the problem of standard BP algorithm, such as slow convergence rate. A comprehensive evaluation model of software quality based on LM-BP neural network is established, which provides a new method and idea for comprehensive evaluation of software quality. The experimental results show that the comprehensive evaluation of software quality based on LM-BP neural network can obtain the comprehensive evaluation result of software quality quickly and accurately, and has good objectivity and practicability.