Variational Inverse Control with Events: A General Framework for Data-Driven Reward Definition

Abstract

The design of a reward function often poses a major practical challenge to real-world applications of reinforcement learning. Approaches such as inverse reinforcement learning attempt to overcome this challenge, but require expert demonstrations, which can be difficult or expensive to obtain in practice. We propose variational inverse control with events (VICE), which generalizes inverse reinforcement learning methods to cases where full demonstrations are not needed, such as when only samples of desired goal states are available. Our method is grounded in an alternative perspective on control and reinforcement learning, where an agent's goal is to maximize the probability that one or more events will happen at some point in the future, rather than maximizing cumulative rewards. We demonstrate the effectiveness of our methods on continuous control tasks, with a focus on high-dimensional observations like images where rewards are hard or even impossible to specify.

arXiv commentFirst two authors contributed equally. Accepted to NIPS. Website: https://sites.google.com/view/inverse-event

References in corpus (7)

• Generative Adversarial Imitation Learning
• Optimal control as a graphical model inference problem
• Bridging the Gap Between Value and Policy Based Reinforcement Learning
• Equivalence Between Policy Gradients and Soft Q-Learning
• Combining policy gradient and Q-learning
• Learning Robust Rewards with Adversarial Inverse Reinforcement Learning
• Inverse Reward Design

Cited by in corpus (13)

• On the Opportunities and Risks of Foundation Models
• Entity Abstraction in Visual Model-Based Reinforcement Learning
• Reward Estimation for Variance Reduction in Deep Reinforcement Learning
• Automating Reinforcement Learning with Example-based Resets
• Example-Driven Model-Based Reinforcement Learning for Solving Long-Horizon Visuomotor Tasks
• Guiding Policies with Language via Meta-Learning
• Lifelong Robotic Reinforcement Learning by Retaining Experiences
• Weakly-Supervised Reinforcement Learning for Controllable Behavior
• Human-guided Robot Behavior Learning: A GAN-assisted Preference-based Reinforcement Learning Approach
• Replacing Rewards with Examples: Example-Based Policy Search via Recursive Classification
• Conservative Data Sharing for Multi-Task Offline Reinforcement Learning
• An Extensible Interactive Interface for Agent Design
• Integration of Imitation Learning using GAIL and Reinforcement Learning using Task-achievement Rewards via Probabilistic Graphical Model