Neural Network Out-of-Distribution Detection for Regression Tasks

Geoff Pleiss, Amauri Souza, Joseph Kim, Boyi Li, Kilian Q. Weinberger

TL;DR: Detect out-of-distribution data on regression neural networks with a generative model of the hidden features

Abstract: Neural network out-of-distribution (OOD) detection aims to identify when a model is unable to generalize to new inputs, either due to covariate shift or anomalous data. Most existing OOD methods only apply to classification tasks, as they assume a discrete set of possible predictions. In this paper, we propose a method for neural network OOD detection that can be applied to regression problems. We demonstrate that the hidden features for in-distribution data can be described by a highly concentrated, low dimensional distribution. Therefore, we can model these in-distribution features with an extremely simple generative model, such as a Gaussian mixture model (GMM) with 4 or fewer components. We demonstrate on several real-world benchmark data sets that GMM-based feature detection achieves state-of-the-art OOD detection results on several regression tasks. Moreover, this approach is simple to implement and computationally efficient.

Keywords: Out-of-distribution, deep learning, regression