Katsuki Ohto
Katsuki Ohto [1]
Katsuki Ohto, a Japanese computer scientist affiliated with the University of Tokyo [2]. As computer game and AI researcher and programmer, he is involved in the development of many game playing programs in the domains of Go, Block Go, Shogi, Dots and Boxes, Trax, Puyo Puyo and Curling simulation.
Programs
His Go playing program Julie which applies deep learning using deep convolutional neural networks [3], participated at the 20th Computer Olympiad, Leiden 2017, winning Bronze in 13x13 Go. His Shogi playing program ShogiNet based on the open source engine Apery [4] applies deep learning as well without search, and became runner-up at the 20th Computer Olympiad, Leiden 2017 behind HGM's Shokidoki.
Selected Publications
- Katsuki Ohto, Tetsuro Tanaka (2016). Application of Monte Carlo Tree Search to Curling AI. 21st Game Programming Workshop
- Katsuki Ohto, Tetsuro Tanaka (2017). A Curling Agent Based on the Monte-Carlo Tree Search Considering the Similarity of the Best Action among Similar States. Advances in Computer Games 15
