Probability estimation for multi-class classification using AdaBoost

作者：

Highlights：

• A probability estimation model is present for classification.

• The model reveals a new view of boosting.

• The model provides a new channel for constructing classifier structure.

• An analytical solution is given for pairwise coupling.

• A robust prediction framework is proposed for pairwise coupling.

摘要

Highlights•A probability estimation model is present for classification.•The model reveals a new view of boosting.•The model provides a new channel for constructing classifier structure.•An analytical solution is given for pairwise coupling.•A robust prediction framework is proposed for pairwise coupling.

论文关键词：AdaBoost,Bayes,Probability estimation,Prediction

论文评审过程：Received 30 October 2013, Revised 4 April 2014, Accepted 11 June 2014, Available online 19 June 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.06.008