Relational Sequential Decision Making

Author(s)

Kaushik Roy, University of South Carolina ColumbiaFollow

Document Type

Paper

Abstract

Markov Decision Processes(MDPs) are the standard for sequential decision making. Comprehensive theory and methods have been developed to deal with solving MDPs in the propositional setting. Real world domains however are naturally represented using objects and relationships. To this effect, relational adaptations of algorithms to solve MDPs have been proposed in recent years. This paper presents a study of these techniques both in the model based and model free setting.

Publication Info

2020.

© Kaushik Roy, 2020

APA Citation

Roy, K. (2020). Relational sequential decision making. [Qualifier defense document].