Model-Free Online Learning in Unknown Sequential Decision Making Problems and Games

03/08/2021
by   Gabriele Farina, et al.
0

Regret minimization has proved to be a versatile tool for tree-form sequential decision making and extensive-form games. In large two-player zero-sum imperfect-information games, modern extensions of counterfactual regret minimization (CFR) are currently the practical state of the art for computing a Nash equilibrium. Most regret-minimization algorithms for tree-form sequential decision making, including CFR, require (i) an exact model of the player's decision nodes, observation nodes, and how they are linked, and (ii) full knowledge, at all times t, about the payoffs – even in parts of the decision space that are not encountered at time t. Recently, there has been growing interest towards relaxing some of those restrictions and making regret minimization applicable to settings for which reinforcement learning methods have traditionally been used – for example, those in which only black-box access to the environment is available. We give the first, to our knowledge, regret-minimization algorithm that guarantees sublinear regret with high probability even when requirement (i) – and thus also (ii) – is dropped. We formalize an online learning setting in which the strategy space is not known to the agent and gets revealed incrementally whenever the agent encounters new decision points. We give an efficient algorithm that achieves O(T^3/4) regret with high probability for that setting, even when the agent faces an adversarial environment. Our experiments show it significantly outperforms the prior algorithms for the problem, which do not have such guarantees. It can be used in any application for which regret minimization is useful: approximating Nash equilibrium or quantal response equilibrium, approximating coarse correlated equilibrium in multi-player games, learning a best response, learning safe opponent exploitation, and online play against an unknown opponent/environment.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/06/2020

Solving imperfect-information games via exponential counterfactual regret minimization

Two agents' decision-making problems can be modeled as the game with two...
research
09/10/2018

Online Convex Optimization for Sequential Decision Processes and Extensive-Form Games

Regret minimization is a powerful tool for solving large-scale extensive...
research
03/08/2021

Bandit Linear Optimization for Sequential Decision Making and Extensive-Form Games

Tree-form sequential decision making (TFSDM) extends classical one-shot ...
research
06/18/2020

DREAM: Deep Regret minimization with Advantage baselines and Model-free learning

We introduce DREAM, a deep reinforcement learning algorithm that finds o...
research
01/13/2023

Decentralized model-free reinforcement learning in stochastic games with average-reward objective

We propose the first model-free algorithm that achieves low regret perfo...
research
10/15/2021

Combining Counterfactual Regret Minimization with Information Gain to Solve Extensive Games with Imperfect Information

Counterfactual regret Minimization (CFR) is an effective algorithm for s...
research
09/15/2020

Finding and Certifying (Near-)Optimal Strategies in Black-Box Extensive-Form Games

Often – for example in war games, strategy video games, and financial si...

Please sign up or login with your details

Forgot password? Click here to reset