Home * People * Tobias Graf

		Tobias Graf, a German computer scientist affiliated with the University of Paderborn. He holds a B.Sc. degree in 2010 on the topic of parallel UCT with high-performance computing ^[1], and a M.Sc. in 2012, elaborating on adaptiv playouts in Monte-Carlo Tree Search applied to Computer Go ^[2]. He is auhor of the Go playing program Abakus, winning 9x9 and 13x13, and runner-up in 9x19 at the 19th Computer Olympiad, and silver and bronze at the 18th Computer Olympiad, and before collaborated with Lars Schaefers on the Go playing program Gomorra. During the Computer and Games conference, 2013 in Yokohama, he lectured on detection of capturing races (Semeai) in Go, co-authored by Lars Schaefers and Marco Platzner ^[3].
Tobias Graf ^[4]

Selected Publications

^[5]

Tobias Graf (2010). Parallelization of the UCT Algorithm on HPC-Clusters. Bachelor's Thesis, University of Paderborn
Tobias Graf, Ulf Lorenz, Marco Platzner, Lars Schaefers (2011). Parallel Monte-Carlo Tree Search for HPC Systems. Euro-Par 2011, pdf
Tobias Graf (2012). Adaptive Playouts in der Monte-Carlo Spielbaumsuche am Anwendungsfall Go. Master's Thesis, University of Paderborn (German)
Tobias Graf, Lars Schaefers, Marco Platzner (2013). On Semeai Detection in Monte-Carlo Go. CG 2013, pdf
Tobias Graf, Marco Platzner (2015). Adaptive Playouts in Monte Carlo Tree Search with Policy Gradient Reinforcement Learning. Advances in Computer Games 14
Tobias Graf, Marco Platzner (2016). Using Deep Convolutional Neural Networks in Monte Carlo Tree Search. CG 2016

External Links

Up one Level