Advances in Game Theory and Algorithmic Economics

The field of game theory and algorithmic economics is witnessing significant developments, with a focus on innovative solutions and advancements in various areas. Researchers are exploring new approaches to characterize Nash equilibria, with a particular emphasis on parallelizable methods and efficient learning algorithms. The study of correlated equilibria in extensive-form games has also led to important findings on complexity and computability. Moreover, the development of new protocols for verifying smooth strategies in bandits and games is enabling more efficient optimization methods. Another notable trend is the investigation of trade-offs between regret and budget violation in bilateral trade, leading to improved regret rates and a better understanding of the underlying mechanisms. Noteworthy papers include:

  • A Parallelizable Approach for Characterizing NE in Zero-Sum Games After a Linear Number of Iterations of Gradient Descent, which proposes a novel method for characterizing Nash equilibria in zero-sum games.
  • Protocols for Verifying Smooth Strategies in Bandits and Games, which introduces new protocols for verifying approximate optimality of strategies in multi-armed bandits and normal-form games.

Sources

Nash Equilibria with Irradical Probabilities

Tie-breaking Agnostic Lower Bound for Fictitious Play

Generalized Quantal Response Equilibrium: Existence and Efficient Learning

The Value Problem for Weighted Timed Games with Two Clocks is Undecidable

Protocols for Verifying Smooth Strategies in Bandits and Games

A Parallelizable Approach for Characterizing NE in Zero-Sum Games After a Linear Number of Iterations of Gradient Descent

Better Regret Rates in Bilateral Trade via Sublinear Budget Violation

On the Complexity of the Optimal Correlated Equilibria in Extensive-Form Games

Built with on top of