Home > Publications . Search All . Browse All . Country . Browse PSC Pubs . PSC Report Series

PSC In The News

RSS Feed icon

H. Luke Shaefer and colleagues argue for a universal child allowance

Hindustan Times points out high value of H-1B visas for US innovation, welfare, and tech firm profits

Novak, Geronimus, Martinez-Cardoso: Threat of deportation harmful to immigrants' health

More News


Heather Ann Thompson wins Pulitzer Prize for book on Attica uprising

Lam explores dimensions of the projected 4 billion increase in world population before 2100

ISR's Nick Prieur wins UMOR award for exceptional contribution to U-M's research mission

How effectively can these nations handle outside investments in health R&D?

More Highlights

Towards min max generalization in reinforcement learning

Publication Abstract

Founteneau, Raphael, Susan A. Murphy, Louis Wehenkel, and Damien Ernst. 2011. "Towards min max generalization in reinforcement learning." In Agents and Artificial Intelligence: International Conference, ICAART 2010, Valencia, Spain, January 2010, Revised Selected Papers, Series: Communications in Computer and Information Science (CCIS) edited by J. Filipe, A. Fred, and B. Sharp. Volume 129, 61-77.

In this paper, we introduce a minmax approach for addressing the generalization problem in Reinforcement Learning. The minmax approach works by determining a sequence of actions that maximizes the worst return that could possibly be obtained considering any dynamics and reward function compatible with the sample of trajectories and some prior knowledge on the environment. We consider the particular case of deterministic Lipschitz continuous environments over continuous state spaces, finite action spaces, and a finite optimization horizon. We discuss the non-triviality of computing an exact solution of the minmax problem even after reformulating it so as to avoid search in function spaces. For addressing this problem, we propose to replace, inside this minmax problem, the search for the worst environment given a sequence of actions by an expression that lower bounds the worst return that can be obtained for a given sequence of actions. This lower bound has a tightness that depends on the sample sparsity. From there, we propose an algorithm of polynomial complexity that returns a sequence of actions leading to the maximization of this lower bound. We give a condition on the sample sparsity ensuring that, for a given initial state, the proposed algorithm produces an optimal sequence of actions in open-loop. Our experiments show that this algorithm can lead to more cautious policies than algorithms combining dynamic programming with function approximators.

Browse | Search : All Pubs | Next