摘要
用神经网络对软件可靠性进行预测比传统的NHPP预测模型预测精度高、拟合度好。但是神经网络的结构是有经验的神经网络专家根据实际经验得来的。而该文提出了用遗传算法去优化神经网络的结构，很好的解决了神经网络的结构问题。用进化的神经网络对软件可靠性的预测提高了预测的精度和准确度。
It is more precise using neural network to predict software reliability than the model of NHPP. The structure of neural network designs by experienced experts. The author uses genetic algorithm to optimize the structure of neural network and solves this problem. The evolutionary neural network can effectively improve the ability of prediction in precision and accurate.
作者
张桂勇
沈元隆
丁晓光
王玲
ZHANG Gui-yong, SHEN Yuan-long, DING Xiao-guang, WANG Ling （College of Optoelectronic Engineering,Nanjing University of Posts ＆ Telecommunications,Nanjing 210003,China）
出处
《电脑知识与技术》
2008年第10期181-183,189,共4页
Computer Knowledge and Technology
关键词
软件可靠性
神经网络
遗传算法
software reliability
neural network
genetic algorithm
