APF+: Boosting adaptive-potential function reinforcement learning methods with a W-shaped network for high-dimensional games

Open in new window