Troy Loeffler1,Sukriti Manna1,Henry Chan1,Rohit Batra1,Subramanian Sankaranarayanan1
Argonne National Laboratory1
Troy Loeffler1,Sukriti Manna1,Henry Chan1,Rohit Batra1,Subramanian Sankaranarayanan1
Argonne National Laboratory1
Monte Carlo Tree Search is a popular algorithm for playing traditional board games such as Go, Chess, Checkers, etc. because of it's ability to opperate and make decisions using limited information about the state of the game. With some adjustments the same AI is capable of being used as powerful optimizer for a wide range of applications. In this presentation we'll explore how this can be used in material science to create new forcefields and perform all manners of useful structure optimization.