An object-oriented representation for efficient reinforcement learning
Carlos Diuk, Andre Cohen, Michael L. Littman · International Conference on Machine Learning · 2008
Rich representations in reinforcement learning have been studied for the purpose of enabling generalization and making learning feasible in large state spaces. We introduce Object-Oriented MDPs (OO-MDPs), a representation based on objects and their interactions, which is a natural way of modeling environments and offers important generalization opportunities. We introduce a learning algorithm for deterministic OO-MDPs and prove a polynomial bound on its sample complexity. We illustrate the performance gains of our representation and algorithm in the well-known Taxi domain, plus a real-life videogame.
Cited by 216 works. doi:10.1145/1390156.1390187
Authors
References
- Dynamic Programming
- Reinforcement Learning: An Introduction
- Artificial intelligence: a modern approach
- A Theory of the Learnable
- Dynamic Programming and Markov Processes.
- Learning to Predict by the Methods of Temporal Differences
- Learning from Delayed Rewards
- Introduction to Reinforcement Learning
- Reinforcement Learning: A Survey
- Learning Quickly When Irrelevant Attributes Abound: A New Linear-Threshold Algorithm