Uncertainty Decomposition in Bayesian Neural Networks with Latent Variables

Abstract

Bayesian neural networks (BNNs) with latent variables are probabilistic models which can automatically identify complex stochastic patterns in the data. We describe and study in these models a decomposition of predictive uncertainty into its epistemic and aleatoric components. First, we show how such a decomposition arises naturally in a Bayesian active learning scenario by following an information theoretic approach. Second, we use a similar decomposition to develop a novel risk sensitive objective for safe reinforcement learning (RL). This objective minimizes the effect of model bias in environments whose stochastic dynamics are described by BNNs with latent variables. Our experiments illustrate the usefulness of the resulting decomposition in active learning and safe RL settings.

arXiv commentThis article is superseded by arXiv:1710.07283

References in corpus (7)

• Estimating Mutual Information
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• Weight Uncertainty in Neural Networks
• Probabilistic Backpropagation for Scalable Learning of Bayesian Neural Networks
• Learning Multimodal Transition Dynamics for Model-Based Reinforcement Learning
• Particle Value Functions
• Introduction to the "Industrial Benchmark"

Cited by in corpus (3)

• Quantification of Uncertainties in Deep Learning-based Environment Perception
• Active Imitation Learning from Multiple Non-Deterministic Teachers: Formulation, Challenges, and Algorithms
• Scalable Bayesian neural networks by layer-wise input augmentation