NeurIPS 2020

Timeseries Anomaly Detection using Temporal Hierarchical One-Class Network

Meta Review

This paper received 3 reviews. The reviewers appreciated the extension of support vector data description (SVDD) to temporal data, with interesting multi-scale feature analysis for anomaly detection in time series. The empirical results are strong. After reading the authors' rebuttal, the reviewers converged on a unanimous accept rating for the paper.