Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness

By Jeremiah Zhe Liu, Zi Lin, Shreyas Padhy, Dustin Tran, Tania Bedrax-Weiss, Balaji Lakshminarayanan. Neural Information Processing Systems, 2020

573 citations recorded by the source corpus.

Bayesian neural networks (BNN) and deep ensembles are principled approaches to estimate the predictive uncertainty of a deep learning model. However their practicality in real-time, industrial-scale applications are limited due to their heavy memory and inference cost. This motivates us to study principled approaches to high-quality uncertainty estimation that require only a single deep neural network (DNN). By formalizing the uncertainty quantification as a minimax learning problem, we first identify input distance awareness, i.e., the model's ability to quantify the distance of a testing example from the training data in the input space, as a necessary condition for a DNN to achieve high-quality (i.e., minimax optimal) uncertainty estimation. We then propose Spectral-normalized Neural Gaussian Process (SNGP), a simple method that improves the distance-awareness ability of modern DNNs, by adding a weight normalization step during training and replacing the output layer with a Gaussian process. On a suite of vision and language understanding tasks and on modern architectures (Wide-ResNet and BERT), SNGP is competitive with deep ensembles in prediction, calibration and out-of-domain detection, and outperforms the other single-model approaches.

Subfields: Artificial Intelligence, Safety, Robustness & Interpretability, Probabilistic & Bayesian Methods

View Paper: https://arxiv.org/abs/2006.10108

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Safety, Robustness & Interpretability

In Probabilistic & Bayesian Methods

Citations in the corpus

Posterior Meta-Replay for Continual Learning (2021, background): “This would naturally lead to an inference strategy as in Lee et al. (2020), where a task-conditioned generative model of X with tractable density is required.”
Deterministic Neural Networks with Appropriate Inductive Biases Capture Epistemic and Aleatoric Uncertainty (2021, methodology): “On the other hand, to quantify the epistemic uncertainty, we use i) the softmax entropy, ii) the softmax density [Liu et al., 2020b] or iii) the GMM feature density (as described in section 4).”
A Simple Fix to Mahalanobis Distance for Improving Near-OOD Detection (2021, methodology): “We choose the following established near-OOD setups: (i) CIFAR-100 vs. CIFAR-10, (ii) CIFAR-10 vs. CIFAR-100, (iii) Genomics OOD benchmark [20] and (iv) CLINC Intent OOD benchmark [11; 13].”
A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges (2020, unknown): “[344] (SNGP: Spectral-normalized Neural GP), Scillitoe et al.”
Exploring the Limits of Out-of-Distribution Detection (2021, methodology): “There has been a lot of progress in detecting OOD inputs including methods based on discriminative models [Hendrycks and Gimpel, 2016, Lee et al., 2018, Liang et al., 2017, Liu et al., 2020] as well as methods based on deep generative models [Nalisnick et al., 2019, Zhang et al., 2020].”
Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty (2020, methodology): “These methods can be broadly classified into two categories: Bayesian approaches [5, 20, 32, 51, 57, 63, 69, 92, 118, 184, 194] and Non-Bayesian, or frequentist, approaches [2, 98, 102, 107, 111, 177].”