Showing 1–1 of 1 results for author: İşeri, M

The Learning Approach to Games

Authors: Melih İşeri, Erhan Bayraktar

Abstract: This work provides a unified framework for exploring games. In existing literature, strategies of players are typically assigned scalar values, and the concept of Nash equilibrium is used to identify compatible strategies. However, this approach lacks the internal structure of a player, thereby failing to accurately model observed behaviors in reality. To address this limitation, we propose to cha… ▽ More This work provides a unified framework for exploring games. In existing literature, strategies of players are typically assigned scalar values, and the concept of Nash equilibrium is used to identify compatible strategies. However, this approach lacks the internal structure of a player, thereby failing to accurately model observed behaviors in reality. To address this limitation, we propose to characterize players by their learning algorithms, and as their estimations intrinsically induce a distribution over strategies, we introduced the notion of equilibrium in terms of characterizing the recurrent behaviors of the learning algorithms. This approach allows for a more nuanced understanding of players, and brings the focus to the challenge of learning that players face. While our explorations in discrete games, mean-field games, and reinforcement learning demonstrate the framework's broad applicability, they also set the stage for future research aimed at specific applications. △ Less

Submitted 28 February, 2025; originally announced March 2025.

Comments: 32 pages, 2 figures. Related repositories are http://github.com/melihiseri/TwoPlayerGame and http://github.com/melihiseri/CartPole_ToyModel

ACM Class: I.2.6; J.4