Simultaneous Abstraction and Equilibrium Finding in Games

نویسندگان

  • Noam Brown
  • Tuomas Sandholm
چکیده

A key challenge in solving extensive-form games is dealing with large, or even infinite, action spaces. In games of imperfect information, the leading approach is to find a Nash equilibrium in a smaller abstract version of the game that includes only a few actions at each decision point, and then map the solution back to the original game. However, it is difficult to know which actions should be included in the abstraction without first solving the game, and it is infeasible to solve the game without first abstracting it. We introduce a method that combines abstraction with equilibrium finding by enabling actions to be added to the abstraction at run time. This allows an agent to begin learning with a coarse abstraction, and then to strategically insert actions at points that the strategy computed in the current abstraction deems important. The algorithm can quickly add actions to the abstraction while provably not having to restart the equilibrium finding. It enables anytime convergence to a Nash equilibrium of the full game even in infinite games. Experiments show it can outperform fixed abstractions at every stage of the run: early on it improves as quickly as equilibrium finding in coarse abstractions, and later it converges to a better solution than does equilibrium finding in fine-grained abstractions.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Abstraction Using Analysis of Subgames

ion Using Analysis of Subgames Anjon Basak and Christopher Kiekintveld 1 The University of Texas at El Paso [email protected], 2 The University of Texas at El Paso [email protected] Abstract. Normal form games are one of the most familiar representations for modeling interactions among multiple agent. However, modeling many realistic interactions between agents results in games that a...

متن کامل

The Path to Equilibrium in Sequential and Simultaneous Games

We study in the laboratory threeand four-player, two-action, dominance solvable games of complete information. We consider sequential and simultaneous versions of games that have the same equilibrium actions, and use mousetracking to determine which payoffs subjects pay attention to. We find slightly more equilibrium choices in sequential than in simultaneous, and an overall good fit of level k...

متن کامل

Simultaneous Decentralized Competitive Supply Chain Network Design under Oligopoly Competition

This paper discusses a problem in which  decentralized supply chains enter the market simultaneously with no existing rival chains, shape the supply chains’ networks, and set wholesale and retail prices in a noncooperative manner. All the chains produce either identical or highly substitutable products. Customer demand is elastic and price-dependent. A three-step algorithm is proposed to solve ...

متن کامل

Dynamic Competitive Supply Chain Network Design with Price Dependent Demand and Huff Utility Function

This paper develops a two-stage model to consider a franchise/franchisee environment in which supply chains are simultaneously entering the untapped market to produce either identical or highly substitutable products and give franchise to franchisees. Customer demand is elastic, price dependent and customer utility function is based on Huff gravity rule model. The supply chains, in the first st...

متن کامل

Nash Equilibrium Strategy for Bi-matrix Games with L-R Fuzzy Payoffs

In this paper, bi-matrix games are investigated based on L-R fuzzy variables. Also, based on the fuzzy max order several models in non-symmetrical L-R fuzzy environment is constructed and the existence condition of Nash equilibrium strategies of the fuzzy bi-matrix games is proposed. At last, based on the Nash equilibrium of crisp parametric bi-matrix games, we obtain the Pareto and weak Pareto...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2015