Temporal Difference Learning for Game 2048 (Demo)
Computer Games and Intelligence (CGI) Lab, NCTU, Taiwan
http://www.aigames.nctu.edu.tw/
Basic framework for training 2048 after-state agent. Including 64-bit bitboard, n-tuple network and isomorphic feature builder, TD(0) training process.
References:
[1] Szubert, Marcin, and Wojciech Jaśkowski. "Temporal difference learning of n-tuple networks for the game 2048."
Computational Intelligence and Games (CIG), 2014 IEEE Conference on. IEEE, 2014.
[2] Wu, I-Chen, et al. "Multi-stage temporal difference learning for 2048."
Technologies and Applications of Artificial Intelligence. Springer International Publishing, 2014. 366-378.
[3] Oka, Kazuto, and Kiminori Matsuzaki. "Systematic selection of n-tuple networks for 2048."
International Conference on Computers and Games. Springer International Publishing, 2016.