Universal Learning of Repeated Matrix Games
Abstract
Description
We study and compare the learning dynamics of two universal learning algorithms, one based on Bayesian learning and the other on prediction with expert advice. Both approaches have strong asymptotic performance guarantees. When confronted with the task of finding good long-term strategies in repeated 2x2 matrix games, they behave quite differently.
16 LaTeX pages, 8 eps figures
16 LaTeX pages, 8 eps figures