Multi-view decision processes: The helper-AI problem
Paper i proceeding, 2017

© 2017 Neural information processing systems foundation. All rights reserved. We consider a two-player sequential game in which agents have the same reward function but may disagree on the transition probabilities of an underlying Markovian model of the world. By committing to play a specific policy, the agent with the correct model can steer the behavior of the other agent, and seek to improve utility. We model this setting as a multi-view decision process, which we use to formally analyze the positive effect of steering policies. Furthermore, we develop an algorithm for computing the agents' achievable joint policy, and we experimentally show that it can lead to a large utility increase when the agents' models diverge.

Författare

Christos Dimitrakakis

Chalmers, Data- och informationsteknik, Datavetenskap

University of Lille

David C. Parkes

Harvard University

Goran Radanovic

Harvard University

Paul Tylkin

Harvard University

Advances in Neural Information Processing Systems

1049-5258 (ISSN)

Vol. 2017-December 5444-5453

31st Annual Conference on Neural Information Processing Systems, NIPS 2017
Long Beach, USA,

Ämneskategorier

Inbäddad systemteknik

Datavetenskap (datalogi)

Datorsystem

Mer information

Senast uppdaterat

2019-05-23