摘要
软件可靠性是软件工程的一个重要的研究课题。软件可靠性模型可以预测软件产品的缺陷达到率。针对软件的缺陷达到率,利用小波神经网络对软件可靠性进行建模。通过理论分析和预测实验表明,该方法比Elman神经网络收敛速度快,逼近效果好,拓宽了软件可靠性模型的研究方法。
The software reliability is an important subject of the software engineering field. Software reliability model can predict defect arrival rate. To predict defect arrival rate,this paper uses the wavelet neural network technology to build software reliability modeling. The demonstration results show that its convergent rate and approaching effectiveness of wavelet neural network is better than Elman neural network. The method is feasible to broaden the software reliability model research methods through theoretical analysis and prediction experiments.
出处
《楚雄师范学院学报》
2016年第3期6-10,共5页
Journal of Chuxiong Normal University
关键词
小波
神经网络
软件可靠性
缺陷达到率
wavelet
neural network
software reliability
defect arrival rate