ML Anthology

Hidden Parameter Markov Decision Processes: A Semiparametric Regression Approach for Discovering Latent Task Parametrizations

Finale Doshi-Velez, George Dimitri Konidaris
IJCAI 2016 pp. 1432-1440
/ijcai/2016/doshivelez2016ijcai-hidden/

Abstract

Control applications often feature tasks with similar, but not identical, dynamics. We introduce the Hidden Parameter Markov Decision Process (HiP-MDP), 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. PDF

PDF IJCAI Semantic Scholar

Cite

Text

Doshi-Velez and Konidaris. "Hidden Parameter Markov Decision Processes: A Semiparametric Regression Approach for Discovering Latent Task Parametrizations." International Joint Conference on Artificial Intelligence, 2016.

Markdown

[Doshi-Velez and Konidaris. "Hidden Parameter Markov Decision Processes: A Semiparametric Regression Approach for Discovering Latent Task Parametrizations." International Joint Conference on Artificial Intelligence, 2016.](https://mlanthology.org/ijcai/2016/doshivelez2016ijcai-hidden/)

BibTeX

@inproceedings{doshivelez2016ijcai-hidden,
  title     = {{Hidden Parameter Markov Decision Processes: A Semiparametric Regression Approach for Discovering Latent Task Parametrizations}},
  author    = {Doshi-Velez, Finale and Konidaris, George Dimitri},
  booktitle = {International Joint Conference on Artificial Intelligence},
  year      = {2016},
  pages     = {1432-1440},
  url       = {https://mlanthology.org/ijcai/2016/doshivelez2016ijcai-hidden/}
}