| Definition | A reinforcement learning environment where agents are trained to play soccer in a primitive 2D simulator |
| No of environments | 1 |
• A reinforcement learning environment where agents are trained to play soccer in a primitive 2D simulator. AI players are trained to play with hard-coded opponents. Developed with API from OpenAI’s Gym.
• Implemented different reinforcement algorithms for training, such as Deep Q learning, PPO(Proximal Policy Optimization) and Advantage Actor Critic (A2C)