Author(s):

Chatterjee, Krishnendu; Elgyutt, Adrian; Novotný, Petr; Rouillé, Owen

Title: 
Expectation optimization with probabilistic guarantees in POMDPs with discountedsum objectives

Affiliation 
IST Austria 
Abstract: 
Partiallyobservable Markov decision processes (POMDPs) with discountedsum payoff are a standard framework to model a wide range of problems related to decision making under uncertainty. Traditionally, the goal has been to obtain policies that optimize the expectation of the discountedsum payoff. A key drawback of the expectation measure is that even low probability events with extreme payoff can significantly affect the expectation, and thus the obtained policies are not necessarily riskaverse. An alternate approach is to optimize the probability that the payoff is above a certain threshold, which allows obtaining riskaverse policies, but ignores optimization of the expectation. We consider the expectation optimization with probabilistic guarantee (EOPG) problem, where the goal is to optimize the expectation ensuring that the payoff is above a given threshold with at least a specified probability. We present several results on the EOPG problem, including the first algorithm to solve it.

Conference Title:

IJCAI: International Joint Conference on Artificial Intelligence

Volume: 
2018

Conference Dates:

July 1319, 2018

Conference Location:

Stockholm, Sweden

ISBN:

9780999241127

Publisher:

IJCAI

Start Page: 
4692

End Page:

4699

DOI: 
10.24963/ijcai.2018/652

Notes: 
This research was supported by the Vienna Science and Technology Fund (WWTF) grant ICT15003; Austrian Science Fund (FWF): S11407N23(RiSE/SHiNE);and an ERC Start Grant (279307:Graph Games).

Open access: 
yes (repository) 