← Back to papers
2018 ICML ICML 2018

Configurable Markov Decision Processes

Abstract

In many real-world problems, there is the possibility to configure, to a limited extent, some environmental parameters to improve the performance of a learning agent. In this paper, we propose a novel framework, Configurable Markov Decision Processes (Conf-MDPs), to model this new type of interaction with the environment. Furthermore, we provide a new learning algorithm, Safe Policy-Model Iteration (SPMI), to jointly and adaptively optimize the policy and the environment configuration. After having introduced our approach and derived some theoretical results, we present the experimental evaluation in two explicative problems to show the benefits of the environment configurability on the performance of the learned policy.

🌉 Interdisciplinary Bridge — Artificial Intelligence and Reinforcement Learning
🧭 Keyword Pioneer — configurable markov decision process
🐝 Cross-Pollinator — Artificial Intelligence, Deep Learning, Machine Learning, Mathematics & Optimization, Natural Language Processing, Reinforcement Learning, Speech & Audio

Authors

Alberto Maria Metelli , Mirco Mutti , Marcello Restelli

Topics

Artificial Intelligence > Core AI > Agent Systems Reinforcement Learning > Methods > Policy Learning Machine Learning > Learning Types > Reinforcement Learning

Keywords

reinforcement learning policy optimization markov decision process joint optimization configurable markov decision process safe policy-model iteration environment configuration
Download PDF

Related papers

Rectify Heterogeneous Models with Semantic Mapping 2018
Bayesian Optimization of Combinatorial Structures 2018
The Well-Tempered Lasso 2018
Approximation Algorithms for Cascading Prediction Models 2018
Classification from Pairwise Similarity and Unlabeled Data 2018