A noise-correction algorithm based on AdaBoost for mislabeled data classification

Xiangyang Liu, Yaping Dai, Guosai Yang, Junjie Ma

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

AdaBoost is one of the most popular algorithm for classification and has been successfully used for text classification, face detection and tracking. However noise sensitivity is regarded as a major disadvantage and previous works show that AdaBoost will be overfitting in dealing with the data sets with noisy data. To improve the noise tolerance of conventional AdaBoost, this paper proposed a Noise-Correction algorithm for Mislabeled Data (NCMD) to find the noisy data and correct it. Further decision stump is selected as the weak learner of the AdaBoost algorithm for classification. Comparison of simulation results between conventional AdaBoost and the method proposed in this paper shows that the proposed algorithm has improved testing accuracy of the data sets with the noisy data.

源语言英语
主期刊名ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications
出版商Fuji Technology Press
ISBN(电子版)9784990534349
出版状态已出版 - 2016
活动7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016 - Beijing, 中国
期限: 3 11月 20166 11月 2016

出版系列

姓名ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications

会议

会议7th International Symposium on Computational Intelligence and Industrial Applications, ISCIIA 2016
国家/地区中国
Beijing
时期3/11/166/11/16

指纹

探究 'A noise-correction algorithm based on AdaBoost for mislabeled data classification' 的科研主题。它们共同构成独一无二的指纹。

引用此

Liu, X., Dai, Y., Yang, G., & Ma, J. (2016). A noise-correction algorithm based on AdaBoost for mislabeled data classification. 在 ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications (ISCIIA 2016 - 7th International Symposium on Computational Intelligence and Industrial Applications). Fuji Technology Press.