Repository logo
Research Data
Publications
Projects
Persons
Organizations
English
Français
Log In(current)
  1. Home
  2. Publications
  3. Article de recherche (journal article)
  4. Reinforcement Learning in the Wild with Maximum Likelihood-based Model Transfer

Reinforcement Learning in the Wild with Maximum Likelihood-based Model Transfer

Author(s)
Hannes Eriksson
Debabrota Basu
Tommy Tram
Mina Alibeigi
Dimitrakakis, Christos  
Chaire de science des données  
Date issued
2023
In
Computing Research Repository (CoRR)
Vol
2302.09273
Subjects
Machine Learning (cs.LG)
Abstract
In this paper, we study the problem of transferring the available Markov Decision Process (MDP) models to learn and plan efficiently in an unknown but similar MDP. We refer to it as \textit{Model Transfer Reinforcement Learning (MTRL)} problem. First, we formulate MTRL for discrete MDPs and Linear Quadratic Regulators (LQRs) with continuous state actions. Then, we propose a generic two-stage algorithm, MLEMTRL, to address the MTRL problem in discrete and continuous settings. In the first stage, MLEMTRL uses a \textit{constrained Maximum Likelihood Estimation (MLE)}-based approach to estimate the target MDP model using a set of known MDP models. In the second stage, using the estimated target MDP model, MLEMTRL deploys a model-based planning algorithm appropriate for the MDP class. Theoretically, we prove worst-case regret bounds for MLEMTRL both in realisable and non-realisable settings. We empirically demonstrate that MLEMTRL allows faster learning in new MDPs than learning from scratch and achieves near-optimal performance depending on the similarity of the available MDPs and the target MDP.
Publication type
journal article
Identifiers
https://libra.unine.ch/handle/20.500.14713/64457
DOI
10.48550/arXiv.2302.09273
File(s)
Loading...
Thumbnail Image
Download
Name

2302.09273.pdf

Type

Main Article

Size

2.31 MB

Format

Adobe PDF

Université de Neuchâtel logo

Service information scientifique & bibliothèques

Rue Emile-Argand 11

2000 Neuchâtel

contact.libra@unine.ch

Service informatique et télématique

Rue Emile-Argand 11

Bâtiment B, rez-de-chaussée

Powered by DSpace-CRIS

libra v2.1.0

© 2025 Université de Neuchâtel

Portal overviewUser guideOpen Access strategyOpen Access directive Research at UniNE Open Access ORCIDWhat's new