A general framework to increase the robustness of model-based change point detection algorithms to outliers and noise.
Date of Submission:
January 23, 2016
The autonomous identification of time-steps where the behavior of a time-series significantly deviates from a predefined model, or time-series change point detection, is an active field of research with notable applications in finance, health, and advertising. One family of time-series change detection algorithms, referred to as ``model-based methods'', although useful for real-time monitoring, are limited when the data are noisy and have outliers. We introduce a new framework that enables any existing method to resilient to these data challenges. We demonstrate the effectiveness of our approach on remote sensing and mobile health data. Our method introduces two new concepts: First a random sampling procedure allows us to overcome outliers, and a matrix-based representation of anomaly scores provides a flexible and intuitive way to identify multiple types of changes and test their significance. We show that our method performs better than several baselines, including application-specific algorithms and provide all data and open-source code to the public.