AE for actions from an expert DQN | using Open AI baselines

modified dqn (Nature architecture)

minor changes!

Added actor-mimic, tensorboard example for video prediction

Added actor-mimic implementation

Merge branch 'master' of https://github.com/manantomar/Reinforcement-Learning

training working for random policies

Create README.md

first commit