GitHub - tzilkha/AlphaNP: NP Problems using AlphaZero and varied state representation using LSTM

tzilkha / AlphaNP Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

NP Problems using AlphaZero and varied state representation using LSTM

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 19 Commits
saved_models		saved_models
AlphaNP_report.pdf		AlphaNP_report.pdf
EVALUATE.py		EVALUATE.py
MCTS.py		MCTS.py
NETS.py		NETS.py
PerfectPolicyTraining.ipynb		PerfectPolicyTraining.ipynb
README.txt		README.txt
SELFPLAY.py		SELFPLAY.py
SelfPlayTraining.ipynb		SelfPlayTraining.ipynb
TSP.py		TSP.py

Repository files navigation

Authors:
Tal Isaac Zilkha (tiz2102)
Mat Hillman (mh3691)

EVALUATE module - 
Utility methods to compare multiple nets.

MCTS module - 
Monte carlo tree search algorithm

NETS module - 
Neural Net classes used for experiments

SELFPLAY module - 
Contains class and methods to train a neural network through monte carlo
assisted self-play to generate samples. 

TSP module - 
Contains logic for the TSP single player game.