| Contention over self-replenishing resources is a common phenomenon which can be simulated via a game of sufficient complexity. The Commons Game provides a framework which allows one to model these types of conflicts.;The obtained results conclusively show that hill-climbing and learning automata are successful at finding optimal strategies for use against random-selection opponents within the confines of the Commons Game. More specifically, we show that learning automata converge to the optimal action in all experimental settings which involve one player using a learning automaton while the remaining players rely on random action-selection.;This work examines the, previously unexplored, subject of application of hill-climbing, particle swarm optimization and a set of learning automata algorithms in order to find the optimal strategies for playing the Commons Game against opponents which use a random action-selection approach. |
