Discovering Blind Spots in Reinforcement Learning

Abstract

Agents trained in simulation may make errors in the real world due to mismatches between training and execution environments. These mistakes can be dangerous and difficult to discover because the agent cannot predict them a priori. We propose using oracle feedback to learn a predictive model of these blind spots to reduce costly errors in real-world applications. We focus on blind spots in reinforcement learning (RL) that occur due to incomplete state representation: The agent does not have the appropriate features to represent the true state of the world and thus cannot distinguish among numerous states. We formalize the problem of discovering blind spots in RL as a noisy supervised learning problem with class imbalance. We learn models to predict blind spots in unseen regions of the state space by combining techniques for label aggregation, calibration, and supervised learning. The models take into consideration noise emerging from different forms of oracle feedback, including demonstrations and corrections. We evaluate our approach on two domains and show that it achieves higher predictive performance than baseline methods, and that the learned model can be used to selectively query an oracle at execution time to prevent errors. We also empirically analyze the biases of various feedback types and how they influence the discovery of blind spots.

arXiv commentTo appear at AAMAS 2018

References in corpus (9)

• Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation
• Deep reinforcement learning from human preferences
• Deep Exploration via Bootstrapped DQN
• Domain Adaptation for Visual Applications: A Comprehensive Survey
• Uncertainty-Aware Reinforcement Learning for Collision Avoidance
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• Successor Features for Transfer in Reinforcement Learning
• Transfer from Simulation to Real World through Learning Deep Inverse Dynamics Model
• Trial without Error: Towards Safe Reinforcement Learning via Human Intervention