A behavioral learning process in games (Q700080): Difference between revisions

The authors study the ``cumulative proportional reinforcement'' rule, with which an agent plays at each period, an action with probability proportional to the cumulative utility obtained by the agent with that action. Asymptotic properties of this learning process are examined. The process considered in a two-player game is also considered and it converges to some known Nash equilibrium or some mixed equilibria.