摘要
PN学习作为一种新型的二元分类器,对结构化的无标签数据有较好的分类性能.软件模块缺陷预测中对无标签样本数据的分类直接影响着预测结果的准确性和可靠性.提出了基于PN学习方法的软件模块缺陷预测模型,结合灰色关联分析方法对实验样本进行降维处理从而提高模型的运算速度.通过实验和分析,证明了本方法的有效性.
PN learning is a new binary classification ,which has better classification performance on the structure of the unlabeled data .Software module defect prediction ,classification of unlabeled data directly affects the accuracy and reliability of the forecast results .T his paper presents a defect prediction model of the PN learning methods-based on software modules ,combined with gray relational analysis of experimen-tal samples to reduce the dimension thereby increasing the speed of operation of the model .Experiment with traditional methods demonstrates the effectiveness of the method .
出处
《西南师范大学学报(自然科学版)》
CAS
CSCD
北大核心
2014年第8期57-61,共5页
Journal of Southwest China Normal University(Natural Science Edition)
基金
河南省教育厅科学技术研究重点项目(13A520148)
河南工程学院博士基金项目(D2012016)
关键词
PN学习
软件模块缺陷预测
灰色关联分析
软件测试
PN learning software modules defect prediction gray relational analysis software testing