In this paper a reward-based abstraction for solving hybrid MDPs is presented. In the proposed method, we gather information about the rewards and the dynamics ...

Markov decision processes (MDPs) have developed as a stan- dard for representing uncertainty in decision-theoretic planning. How- ever, MDPs require an explicit ...

This paper proposes a novel and practical model-based learning approach with iterative refinement for solving continuous (and hybrid) Markov decision processes.

This paper proposes a reward-based abstraction for solving hybrid MDPs and presents the results in terms of the models learned and their solutions for ...

Abstract. Markov decision processes (MDPs) have developed as a stan- dard for representing uncertainty in decision-theoretic planning. How-.

In this paper a reward-based abstraction for solving hybrid MDPs is presented. In the proposed method, we gather information about the rewards and the dynamics ...

Jul 11, 2012 · We present a new linear program approximation method that exploits the structure of the hybrid MDP and lets us compute approximate value functions more ...

Approximate Linear Programming (ALP) for solving large factored MDPs with continuous and hybrid state components. Partially observable Markov decision processes ...

Introduction. Because of communication limitations, remote spacecraft and rovers need the ability to operate autonomously. For.

In this section we introduce Parameterized Hybrid Markov. Decision Processes (PHMDPs). 3.1 Definition. A parameterized hybrid Markov Decision Process (PH-. MDP) ...