基于人工神经网络的软件失效预测模型研究  

Research of Software Faults Prediction Model Based on Artificial Neural Network

作　　者：王琪[1] 于波 朱杰[1] 

机构地区：[1]上海交通大学电子工程系,上海200030 [2]朗讯科技光网络有限公司,上海200233

出　　处：《计算机仿真》2005年第3期159-161,共3页Computer Simulation

摘　　要：在软件开发的早期预测有失效倾向的软件模块,能够极大的提高软件的质量。软件失效预测中的一个普遍的问题是数据中存在噪声,而神经网络具有鲁棒性并对噪声有很强的抑制能力。该文介绍了一种基于人工神经网络的软件失效预测模型,给出了基于反向传播算法的多层前向网络的网络结构。用这种方法对朗讯光网络有限公司开发的SDH通信软件进行了分析,并得到了较高的预测准确率。通过采集通信软件的不同发布版本的测试历史数据,讨论了训练集数据的选择与预测精度之间的关系。Predicting the fault-prone software module early in the software development can improve the software quality significantly. A common problem in software faults prediction is the presence of noise in the data. Neural Networks are robust and have a good noise tolerance. This paper presents a kind of software faults prediction model based on artificial neural network and the structure of the feed-forward multi-layer network with backpropagation learning algorithm. We use this approach to analyse the SDH telecommunication software which was developed by Lucent Technology Optical Network Company and get high predicting accuracy. We select training sets and test sets in many different software releases and discuss the relation between training set and the prediction accuracy.

关 键 词：软件失效预测 工神经网络 反向传播算法 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]