learning environment; stopping rule; learning success; probability; stochastic models; multi -agents; stochastic game