この資料をSMS送信: Exploration and value function factorisation in single and multi-agent reinforcement learning