BioInspired

So this is the repo made for learning to play Ms.Pacman with Pacman.

Ran on collab, manually choose file path to load the model.

Networks:

1. Q-Network
2. Dueling Q-Network

Transfer Learning:

1. Non Progressive
2. Proggresive Neural Networks

Agent options:

1. Q-Learning Type : Deep & Double (x2)
2. Buffer Type : Replay & PER (x2)
3. Exploration : Epsilon-Greedy (& ICM) (x2)