Strategy Learning
ByHamidou Tembine
Pages 84

This chapter presents basic learning schemes in games. There are two main parts. The first part focuses on strategy learning in games under perfect monitoring of past actions. This class of schemes is referred to partially distributed learning algorithms or semi-distributed schemes, which basically introduces learning mechanisms that use information about the other players (information about their past actions or their past payoffs via some signals or messages, public or private), and in this context, learning different solution concepts such as Nash equilibrium, and bargaining solution in both pure and mixed strategies are introduced. Convergence, nonconvergence properties and features of learning algorithms are thoroughly investigated.