Robust Spectral Filtering and Anomaly Detection
This work addresses anomaly detection for operators of linear dynamical systems, but it appears incremental as it builds on existing robust filtering techniques.
The paper tackles the problem of predicting outputs and identifying noise in linear dynamical systems where outputs are randomly replaced by noise, presenting a robust method with proven statistical guarantees.
We consider a setting, where the output of a linear dynamical system (LDS) is, with an unknown but fixed probability, replaced by noise. There, we present a robust method for the prediction of the outputs of the LDS and identification of the samples of noise, and prove guarantees on its statistical performance. One application lies in anomaly detection: the samples of noise, unlikely to have been generated by the dynamics, can be flagged to operators of the system for further study.