Dropout Inference in Bayesian Neural Networks with Alpha-divergences

Abstract

To obtain uncertainty estimates with real-world Bayesian deep learning models, practical inference approximations are needed. Dropout variational inference (VI) for example has been used for machine vision and medical applications, but VI can severely underestimates model uncertainty. Alpha-divergences are alternative divergences to VI's KL objective, which are able to avoid VI's uncertainty underestimation. But these are hard to use in practice: existing techniques can only use Gaussian approximating distributions, and require existing models to be changed radically, thus are of limited use for practitioners. We propose a re-parametrisation of the alpha-divergence objectives, deriving a simple inference technique which, together with dropout, can be easily implemented with existing models by simply changing the loss of the model. We demonstrate improved uncertainty estimates and accuracy compared to VI in dropout networks. We study our model's epistemic uncertainty far away from the data using adversarial images, showing that these can be distinguished from non-adversarial images by examining our model's uncertainty.

References in corpus (4)

• Explaining and Harnessing Adversarial Examples
• Improving neural networks by preventing co-adaptation of feature detectors
• Expectation Propagation for approximate Bayesian inference
• Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networks

Cited by in corpus (18)

• Deep and Confident Prediction for Time Series at Uber
• Adversarial Examples, Uncertainty, and Transfer Testing Robustness in Gaussian Process Hybrid Deep Networks
• Model-Predictive Policy Learning with Uncertainty Regularization for Driving in Dense Traffic
• Understanding the Decision Boundary of Deep Neural Networks: An Empirical Study
• Active Learning for Deep Detection Neural Networks
• From Deterministic to Generative: Multi-Modal Stochastic RNNs for Video Captioning
• The Hidden Uncertainty in a Neural Networks Activations
• Bayesian Neural Networks
• On the use of uncertainty in classifying Aedes Albopictus mosquitoes
• Stochastic Maximum Likelihood Optimization via Hypernetworks
• Bayesian Optimization Meets Laplace Approximation for Robotic Introspection
• Uncertainty-driven ensembles of deep architectures for multiclass classification. Application to COVID-19 diagnosis in chest X-ray images
• Reliable Uncertainties for Bayesian Neural Networks using Alpha-divergences
• Requirements for Developing Robust Neural Networks
• SINVAD: Search-based Image Space Navigation for DNN Image Classifier Test Input Generation
• Quantal synaptic dilution enhances sparse encoding and dropout regularisation in deep networks
• Robustness Against Outliers For Deep Neural Networks By Gradient Conjugate Priors
• Contributions to Large Scale Bayesian Inference and Adversarial Machine Learning