摘要
BP神经网络、RBF神经网络、支持向量机(SVM)和集成学习是目前应用最为广泛的四种机器学习方法。将这四种常用的机器学习方法分别应用于人脸识别,并利用ORL人脸图像库对各学习方法性能进行了测试和评估。测试结果表明SVM和集成学习在实验中取得了较好的性能,最适合用于人脸识别中特征分类器。
BP neural network,RBF neural network,Support Vector Machines(SVM),and ensemble learning are four widely-used machine learning methods at present.In this paper,these four widely-used machine learning methods are applied to face recognition,and then the ORL database is selected to test and evaluate each learning method.Experimental results demonstrate that SVM and ensemble learning methods have achieved good performance in the experiment and are most suitable for feature classifier in face recognition.
出处
《计算机工程与应用》
CSCD
北大核心
2009年第4期169-172,共4页
Computer Engineering and Applications
基金
国家自然科学基金项目No.60572128
安徽省人才发展基金No.2005Z029
安徽大学创新团队项目基金~~
关键词
人脸识别
机器学习
反向传播神经网络
径向基函数神经网络
支持向量化
集成学习
比较
face recognition
machine learning
Back Propagation(BP) neural network
Radial Basis Function(RBF) neural net-work
Support Vector Machines(SVM)
ensemble learning
comparison