Towards explaining anomalies: A deep Taylor decomposition of one-class models
作者：
Highlights：
• We enhance the prediction of anomalies (as given by a kernel one-class SVM) by explaining them in terms of input features.
• The method is based on a reformulation of the one-class SVM as a neural network, the structure of which is better suited to the task of explanation.
• Explanations are obtained via a deep Taylor decomposition, which propagates the prediction backward in the neural network towards the input features.
• Application of our method to image data highlights pixel-level anomalies that can be missed by a simple visual inspection.
摘要
•We enhance the prediction of anomalies (as given by a kernel one-class SVM) by explaining them in terms of input features.•The method is based on a reformulation of the one-class SVM as a neural network, the structure of which is better suited to the task of explanation.•Explanations are obtained via a deep Taylor decomposition, which propagates the prediction backward in the neural network towards the input features.•Application of our method to image data highlights pixel-level anomalies that can be missed by a simple visual inspection.
论文关键词：Outlier detection,Explainable machine learning,Deep Taylor decomposition,Kernel machines,Unsupervised learning
论文评审过程：Received 6 December 2018, Revised 3 December 2019, Accepted 8 January 2020, Available online 9 January 2020, Version of Record 21 January 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107198
