Learning to Play Soccer From Scratch: Sample-Efficient Emergent Coordination Through Curriculum-Learning and Competition
Pavan Samtani,Francisco Leiva,Javier Ruiz-del-Solar,Pavan Samtani,Francisco Leiva,Javier Ruiz-del-Solar
This work proposes a scheme that allows learning complex multi-agent behaviors in a sample efficient manner, applied to 2v2 soccer. The problem is formulated as a Markov game, and solved using deep reinforcement learning. We propose a basic multi-agent extension of TD3 for learning the policy of each player, in a decentralized manner. To ease learning, the task of 2v2 soccer is divided in three st...
