Work

Multi-Objective MDPs with Conditional Lexicographic Reward Preferences

Year: 2015

Type: conference-paper

Abstract: Sequential decision problems that involve multiple objectives are prevalent. Consider for example a driver of a semi-autonomous car who may want to optimize competing objectives such as travel time and the effort associated with manual driving. We introduce a rich model called Lexicographic MDP (LMDP) and a corresponding planning algorithm called LVI that generalize previous work by allowing for conditional lexicographic preferences with slack. We analyze the convergence characteristics of LVI and establish its game theoretic properties. The performance of LVI in practice is tested within a realistic benchmark problem in the domain of semi-autonomous driving. Finally, we demonstrate how GPU-based optimization can improve the scalability of LVI and other value iteration algorithms for MDPs.

Source: Proceedings of the AAAI Conference on Artificial Intelligence

DOI: 10.1609/aaai.v29i1.9647

Authors: Kyle Wray, Shlomo Zilberstein, Abdel-Illah Mouaddib

Institutions: University of Massachusetts Amherst, Université de Caen Normandie

Language: English

Topic: Reinforcement Learning in Robotics

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: diamond

Funder: National Science Foundation

Award: RI: Medium: Computational Models, Interaction Mechanisms, and Planning Algorithms for Semi-Autonomous Systems