Adaptive learning in large populations

We consider the adaptive learning rule of Harley (J Theor Biol 89:611-633, 1981) for behavior selection in symmetric conflict games in large populations. This rule uses organisms' past, accumulated rewards as the predictor for future behavior, and can be traced in many life forms from bacteria to humans. We derive a partial differential equation for the distribution of agents in the space of stimuli to select a particular strategy which describes the evolution of learning in heterogeneous populations. We analyze the solutions of the PDE model for symmetric [Formula: see text] games. It is found that in games with small residual stimuli, adaptive learning rules with larger memory factor converge faster to the optimal outcome.