- Original Pdf: pdf
- Keywords: Neural Architecture Search, Bayesian ensembling, out-of-distribution detection, uncertainty quantification, density estimation
- TL;DR: We propose an architecture search method to identify a distribution of architectures and use it to construct a Bayesian ensemble for outlier detection.
- Abstract: Machine learning systems often encounter Out-of-Distribution (OoD) errors when dealing with testing data coming from a different distribution from the one used for training. With their growing use in critical applications, it becomes important to develop systems that are able to accurately quantify its predictive uncertainty and screen out these anomalous inputs. However, unlike standard learning tasks, there is currently no well established guiding principle for designing architectures that can accurately quantify uncertainty. Moreover, commonly used OoD detection approaches are prone to errors and even sometimes assign higher likelihoods to OoD samples. To address these problems, we first seek to identify guiding principles for designing uncertainty-aware architectures, by proposing Neural Architecture Distribution Search (NADS). Unlike standard neural architecture search methods which seek for a single best performing architecture, NADS searches for a distribution of architectures that perform well on a given task, allowing us to identify building blocks common among all uncertainty aware architectures. With this formulation, we are able to optimize a stochastic outlier detection objective and construct an ensemble of models to perform OoD detection. We perform multiple OoD detection experiments and observe that our NADS performs favorably compared to state-of-the-art OoD detection methods.
- Code: https://anonymous.4open.science/r/e514158f-b2f9-4b61-a0eb-d05fdb250bd9/