Sequential Adversarial Anomaly Detection for One-Class Event Data
This addresses the challenge of detecting anomalous sequences in domains like fraud detection, but it is incremental as it builds on existing adversarial and marked point process methods.
The paper tackled the problem of sequential anomaly detection in a one-class setting where only anomalous sequences are available, by proposing an adversarial sequential detector that solves a minimax problem to optimize against worst-case sequences, and demonstrated good performance in simulations and on a large-scale credit card fraud dataset.
We consider the sequential anomaly detection problem in the one-class setting when only the anomalous sequences are available and propose an adversarial sequential detector by solving a minimax problem to find an optimal detector against the worst-case sequences from a generator. The generator captures the dependence in sequential events using the marked point process model. The detector sequentially evaluates the likelihood of a test sequence and compares it with a time-varying threshold, also learned from data through the minimax problem. We demonstrate our proposed method's good performance using numerical experiments on simulations and proprietary large-scale credit card fraud datasets. The proposed method can generally apply to detecting anomalous sequences.