Hidden parameter markov decision processes: A semiparametric regression approach for discovering latent task parametrizations

Published

Conference Paper

Control applications often feature tasks with similar, but not identical, dynamics. We introduce the Hidden Parameter Markov Decision Process (HiPMDP), a framework that parametrizes a family of related dynamical systems with a low-dimensional set of latent factors, and introduce a semiparametric regression approach for learning its structure from data. We show that a learned HiP-MDP rapidly identifies the dynamics of new task instances in several settings, flexibly adapting to task variation.

Duke Authors

Cited Authors

  • Doshi-Velez, F; Konidaris, G

Published Date

  • January 1, 2016

Published In

Volume / Issue

  • 2016-January /

Start / End Page

  • 1432 - 1440

International Standard Serial Number (ISSN)

  • 1045-0823

Citation Source

  • Scopus