Doubly Robust Bayesian Inference for Non-Stationary Streaming Data with -Divergences

Abstract

We present the very first robust Bayesian Online Changepoint Detection algorithm through General Bayesian Inference (GBI) with -divergences. The resulting inference procedure is doubly robust for both the predictive and the changepoint (CP) posterior, with linear time and constant space complexity. We provide a construction for exponential models and demonstrate it on the Bayesian Linear Regression model. In so doing, we make two additional contributions: Firstly, we make GBI scalable using Structural Variational approximations that are exact as 0. Secondly, we give a principled way of choosing the divergence parameter by minimizing expected predictive loss on-line. Reducing False Discovery Rates of from up to 99\% to 0\% on real world data, this offers the state of the art.

Tasks

Reproductions