完美密度模型不能保证异常检测

Dec, 2020

Perfect density models cannot guarantee anomaly detection

Charline Le Lan, Laurent Dinh

TL;DR通过重新参数化的视角深入研究分布密度的行为，揭示在现有应用中赋予异常点密度的经验值与应用期望相冲突，提出利用这些似然函数进行异常检测依赖强的暗含假设，明确表述这些假设才能实现可靠异常检测。

Abstract

Thanks to the tractability of their likelihood, some deep generative models show promise for seemingly straightforward but important applications like anomaly detection, →