Anomaly detection via Gumbel Noise Score Matching
Ahsan Mahmooda, Junier B. Olivaa, Martin Andreas Stynera✉
- aDepartment of Computer Science, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States
- ✉Corresponding author
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
We propose Gumbel Noise Score Matching (GNSM), a novel unsupervised method to detect anomalies in categorical data. GNSM accomplishes this by estimating the scores, i.e., the gradients of log likelihoods w.r.t. inputs, of continuously relaxed categorical distributions. We test our method on a suite of anomaly detection tabular datasets. GNSM achieves a consistently high performance across all experiments. We further demonstrate the flexibility of GNSM by applying it to image data where the model is tasked to detect poor segmentation predictions. Images ranked anomalous by GNSM show clear segmentation failures, with the anomaly scores strongly correlating with segmentation metrics computed on ground-truth. We outline the score matching training objective utilized by GNSM and provide an open-source implementation of our work.
TL;DRGumbel Noise Score Matching is proposed, a novel unsupervised method to detect anomalies in categorical data by estimating the scores, i.e., the gradients of log likelihoods w.r.t. inputs, of continuously relaxed categorical distributions. (Semantic Scholar)