Deep Reinforcement learning project, MVA 2019-2020
Since I ran my notebooks on google colab, I was not able to keep the video on my ipynb file. That's why i present here an example of game from, trained agents with a fully connected network, a CNN network and a CNN work in the exploring environment.
Fully connected:
CNN :
Exploration + CNN :