An Analysis of Model-Based Reinforcement Learning From Abstracted Observations

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Many methods for Model-based Reinforcement learning (MBRL) in Markov decision processes (MDPs) provide guarantees for both the accuracy of the model they can deliver and the learning efficiency. At the same time, state abstraction techniques allow for a reduction of the size of an MDP while maintaining a bounded loss with respect to the original problem. Therefore, it may come as a surprise that no such guarantees are available when combining both techniques, i.e., where MBRL merely observes abstract states. Our theoretical analysis shows that abstraction can introduce a dependence between samples collected online (e.g., in the real world). That means that, without taking this dependence into
account, results for MBRL do not directly extend to this setting. Our result shows that we can use concentration inequalities for martingales to overcome this problem. This result makes it possible to extend the guarantees of existing MBRL algorithms to the setting with abstraction. We illustrate this by combining R-MAX, a prototypical MBRL algorithm, with abstraction, thus producing the first performance guarantees for model-based ‘RL from Abstracted Observations’: model-based reinforcement learning with an abstract model.
Original languageEnglish
Number of pages36
JournalTransactions on Machine Learning Research
Publication statusE-pub ahead of print - 2023

Keywords

  • Reinforcement learning (RL)
  • Model-based reinforcement learning
  • abstraction
  • State Abstraction

Fingerprint

Dive into the research topics of 'An Analysis of Model-Based Reinforcement Learning From Abstracted Observations'. Together they form a unique fingerprint.

Cite this