Chris J. Maddison

From Chessprogramming wiki
Jump to: navigation, search

Home * People * Chris J. Maddison

Chris J. Maddison [1]

Chris J. Maddison,
a Canadian computer scientist affiliated with the University of Toronto. His reaearch interests include machine learning, neural networks, deep learning, inference, Monte Carlo methods, and how can we simulate complex models of the world [2].

DCNNs in Go

In their 2014 paper Move Evaluation in Go Using Deep Convolutional Neural Networks [3], Chris J. Maddison, Aja Huang, Ilya Sutskever, and David Silver investigate whether deep convolutional neural networks can be used to directly represent and learn a move evaluation function for the game of Go. They train a large 12-layer convolutional neural network by supervised learning from a database of human professional games. The network correctly predicted the expert move in 55% of positions, equalling the accuracy of a 6 dan human player. When the trained convolutional network was used directly to play games of Go, without any search, it beat the traditional search program Gnu Go in 97% of games, and matched the performance of a state-of-the-art Monte-Carlo Tree Search that simulates a million positions per move [4].

Selected Publications


External Links


Up one level