Error sensitivity analysis of Delta divergence - a novel measure for classifier incongruence detection

作者：

Highlights：

• Theoretical and experimental error sensitivity analysis of Delta divergence as a measure of classifier incongruence.

• Practical guidelines for selecting classifier incongruence thresholds in practice.

• Experimental analysis of Delta divergence under varying noise and classier a posteriori probability distributions.

摘要

•Theoretical and experimental error sensitivity analysis of Delta divergence as a measure of classifier incongruence.•Practical guidelines for selecting classifier incongruence thresholds in practice.•Experimental analysis of Delta divergence under varying noise and classier a posteriori probability distributions.

论文关键词：Anomaly detection,Classifier decision incongruence,Bayesian surprise

论文评审过程：Received 15 February 2017, Revised 24 October 2017, Accepted 30 November 2017, Available online 5 December 2017, Version of Record 27 December 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.11.031