摘要
文章将支持向量机理论引入到软件可靠性早期预测领域中来,提出了基于支持向量机的软件可靠性早期预测模型;通过对比仿真实验,证实了该模型同传统模型相比,具有预测精度高、泛化能力强及对样本数量的依赖程度低的特点。
The support vector machine（SVM） theory is introduced into the field of early software reliability prediction, and a model of early software reliability prediction based on the SVM is put forward. Simulation results show that compared with classic models, the new model has better prediction precision, better generalization ability and lower dependence on the number of samples.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2007年第7期859-863,共5页
Journal of Hefei University of Technology：Natural Science
关键词
软件可靠性
早期预测
支持向量机
software reliability
early prediction
support vector machine（SVM）
