RL agent that learns to play Ping Pong

Implemented the system using policy gradient by creating two layer neural network and applied RMSProp.

Analysis

Rewards before training:

Rewards after training:

Open Gym: loads game environment
Pickle: stores the parameter values and rewards

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
PingPong.ipynb		PingPong.ipynb
README.md		README.md
demo.mp4		demo.mp4
endRewards.png		endRewards.png
pingPong.png		pingPong.png
rewards.p		rewards.p
save.p		save.p
startRewards.png		startRewards.png