TY - GEN
T1 - A scalable neural network architecture for board games
AU - Schaul, Tom
AU - Schmidhuber, Jürgen
N1 - Generated from Scopus record by KAUST IRTS on 2022-09-14
PY - 2008/12/1
Y1 - 2008/12/1
N2 - This paper proposes to use Multi-dimensional Recurrent Neural Networks (MDRNNs) as a way to overcome one of the key problems in flexible-size board games: scalability. We show why this architecture is well suited to the domain and how it can be successfully trained to play those games, even without any domain-specific knowledge. We find that performance on small boards correlates well with perform~nce on large ones, and that this property holds for networks tramedby either evolution or coevolution. ©2008 IEEE.
AB - This paper proposes to use Multi-dimensional Recurrent Neural Networks (MDRNNs) as a way to overcome one of the key problems in flexible-size board games: scalability. We show why this architecture is well suited to the domain and how it can be successfully trained to play those games, even without any domain-specific knowledge. We find that performance on small boards correlates well with perform~nce on large ones, and that this property holds for networks tramedby either evolution or coevolution. ©2008 IEEE.
UR - http://ieeexplore.ieee.org/document/5035662/
UR - http://www.scopus.com/inward/record.url?scp=70349280482&partnerID=8YFLogxK
U2 - 10.1109/CIG.2008.5035662
DO - 10.1109/CIG.2008.5035662
M3 - Conference contribution
SN - 9781424429745
SP - 357
EP - 364
BT - 2008 IEEE Symposium on Computational Intelligence and Games, CIG 2008
ER -