Optimizing area under the ROC curve using semi-supervised learning

作者：

Highlights：

• Optimizing area under the ROC curve using semi-supervised learning.

• A large margin maximization semi-supervised learning framework for AUC maximization.

• Closed-form solution based on semi-definite programming.

• Superior performance on 34 UCI machine learning datasets determined by power analysis.

• Showed efficacy on a CT colonography dataset for colonic polyp classification.

摘要

Highlights•Optimizing area under the ROC curve using semi-supervised learning.•A large margin maximization semi-supervised learning framework for AUC maximization.•Closed-form solution based on semi-definite programming.•Superior performance on 34 UCI machine learning datasets determined by power analysis.•Showed efficacy on a CT colonography dataset for colonic polyp classification.

论文关键词：Receiver operating characteristic,AUC,Semi-supervised learning,Transfer learning,Semidefinite programming,RankBoost,SVMROC,SSLROC

论文评审过程：Received 27 June 2013, Revised 4 July 2014, Accepted 28 July 2014, Available online 6 August 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.07.025