全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

一种新的改进AdaBoost弱分类器训练算法

DOI: 10.11834/jig.20091137

Keywords: 弱分类器,AdaBoost算法,强分类器,错分率

Full-Text   Cite this paper   Add to My Lib

Abstract:

AdaBoost是机器学习中比较流行的分类算法。通过研究弱分类器的特性,提出了两种新的弱分类器的阈值和偏置计算方法,二者可以使弱分类器识别率大于50%,从而保证在弱分类器达到一定数目的情况下,AdaBoost训练收敛。对两种阈值和偏置计算方法的仿真实验结果表明,在错分率降可接受的范围内,二者均使用较少的弱分类器便可获得高识别率的强分类器。

References

[1]  Shan S,Yang P,Chen X,AdaBoost gabor fisher classifier for face recognition,北京,2005.
[2]  Lin Wei-Chan,Oakes M,Tait J,Real AdaBoast for large vocabulary image classification,London,Unit Kingdom,2008.
[3]  Viola P,Jones M,Fast and robust classification using asymmetric AdaBooat and a detector cascade,Cambridge,MA,USA:MIT Press,2002.
[4]  Kim Jeong-hyun,Park Jong-hyun,Kang Dong-joong,Method to improve the performance of the AdaBoost algorithm using gaussian probabilitydistribution.Control,Seoul,Korea (South),2008.
[5]  Dai Sheng-Yang,Zhang Yu-Jin,AdaBoost in region-based image retrieval,Montreal,Canada,2004.
[6]  Leshem G,Ritov Y,Traffic flow prediction using AdaBoast algorithm with random forests as a weak learner,Bangkok,Thailand,2007.
[7]  Yin X C,Liu C P,Han Z,Feature combination using boosting,Pattern Recognition Letters,2005(14).
[8]  李闯,丁晓青,吴佑寿,一种改进的AdaBoost算法--AD AdaBooat,计算机学报,2007(01).

Full-Text

comments powered by Disqus

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133

WeChat 1538708413