The main objective of transfer in reinforcement learning is to reduce the complexity of learning the solution of a target task by effectively reusing the knowledge retained from solving a set of source tasks. In this paper, we introduce a novel algorithm that transfers samples (i.e., tuples <s, a, s', r>) from source to target tasks. Under the assumption that tasks have similar transition models and reward functions, we propose a method to select samples from the source tasks that are mostly similar to the target task, and, then, to use them as input for batch reinforcement-learning algorithms. As a result, the number of samples an agent needs to collect from the target task to learn its solution is reduced. We empirically show that, following the proposed approach, the transfer of samples is effective in reducing the learning complexity, even when some source tasks are significantly different from the target task.

Transfer of samples in batch reinforcement learning

RESTELLI, MARCELLO;BONARINI, ANDREA
2008-01-01

Abstract

The main objective of transfer in reinforcement learning is to reduce the complexity of learning the solution of a target task by effectively reusing the knowledge retained from solving a set of source tasks. In this paper, we introduce a novel algorithm that transfers samples (i.e., tuples ) from source to target tasks. Under the assumption that tasks have similar transition models and reward functions, we propose a method to select samples from the source tasks that are mostly similar to the target task, and, then, to use them as input for batch reinforcement-learning algorithms. As a result, the number of samples an agent needs to collect from the target task to learn its solution is reduced. We empirically show that, following the proposed approach, the transfer of samples is effective in reducing the learning complexity, even when some source tasks are significantly different from the target task.
2008
Proceedings of the 25th Annual International Conference on Machine Learning (ICML 2008)
9781605582054
File in questo prodotto:
File Dimensione Formato  
479.pdf

Accesso riservato

: Post-Print (DRAFT o Author’s Accepted Manuscript-AAM)
Dimensione 360.81 kB
Formato Adobe PDF
360.81 kB Adobe PDF   Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/549209
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 120
  • ???jsp.display-item.citation.isi??? 0
social impact