[2016-04-08] Prof. Shun-Shii Lin, National Taiwan Normal University, "From Ko fight, Computer Olympiad gold, to AlphaGo beating the world Go champion" (從劫爭、奪奧金、到AlphaGo贏棋王之路)

Poster:Post date:2016-03-29
Title: From Ko fight, Computer Olympiad gold, to AlphaGo beating the world Go champion (從劫爭、奪奧金、到AlphaGo贏棋王之路)
Date: 2016-04-08  2:20pm-3:30pm
Location: R103, CSIE
Speaker: Prof. Shun-Shii Lin (林順喜), Department of Computer Science and Information Engineering, National Taiwan Normal University
Hosted by: Prof. Kun-Mao Chao


It is an exciting moment that AlphaGo beats the world Go champion and it represents a historical moment for the advancement of AI. Hence, I would like to talk something about the road from Ko fight, Computer Olympiad gold, to AlphaGo beating the world Go champion.
Ko plays a very important role in Go, but most earlier-developed computer Go programs still cannot handle ko fights. In 2001, we (I and Aja Huang) started the first study to deal with this difficult problem. Utilizing the principle of game theory, we obtain the best strategies for the Go programs to gain maximum or loss minimum profit when dealing with the Ko fights.
In 2011, we (Aja Huang, Rémi Coulom and I) proposed some new heuristics of Monte Carlo Tree Search focused on the successful application of Simulation Balancing and various time management schemes for Go. These results let our Go program ERICA win the gold medal in the 19×19 Go tournament at the 2010 Computer Olympiad, Japan. Many techniques like Monte Carlo Tree Search were still used in AlphaGo.
Due to these outstanding results, a road was occasionally built at NTNU in 10 years to advance the birth of AlphaGo to beat the human world Go champion.

在2011年,我們(黃士傑、Rémi Coulom和我)針對「蒙地卡羅樹搜尋」提出一些新的啟發式演算法,成功的將「模擬平衡化」(Simulation Balancing)應用到圍棋。另提出各種不同之時間控制的方法。所有的實驗都是執行在我們的圍棋程式 ERICA,而 ERICA 正是得益於這些啟發式演算法與實驗結果,成功取得了2010 年在日本舉辦的電腦奧林匹亞的19路圍棋金牌。許多技術如「蒙地卡羅樹搜尋」仍被用於AlphaGo。

Shun-Shii Lin (林順喜) received the B.S. degree in Computer Engineering from the National Chiao-Tung University in 1981 and the M.S. and Ph.D. degrees in Computer Science and Information Engineering from the National Taiwan University in 1985 and 1990, respectively. From August 1986 to July 2001, he was on the faculty of the Department of Information and Computer Education of the National Taiwan Normal University. He was also the head of the department from August 1994 to July 1996. He was a Visiting Scholar in the Department of Computer Science, University of Illinois at Urbana-Champaign from January 1993 to January 1994. Starting August 2001, he is a professor and also serves as the director of the Graduate Institute of Computer Science and Information Engineering of the National Taiwan Normal University from August 2001 to July 2005. Now he is a professor at the Department of Computer Science and Information Engineering, National Taiwan Normal University. Dr. Lin is the inventors of the Outer-Open Gomoku and NTNU-Master input method, a member of the Institute of Information and Computing Machinery and the Phi Tau Phi Scholastic Honor Society. His research interests include design/analysis of algorithms and artificial intelligence. Till 2015, he has advised his students to win 17 gold medals in the domestic/international Computer Game Tournaments. He has advised 10 Master students to win the Outstanding Master's thesis Awards. He has also advised 5 Ph.D graduated students. Among them, 3 have won the Outstanding Ph.D dissertation Awards. His most famous student is Aja Huang (黃士傑) who is one of the key developers behind AlphaGo that beats the world human Go champion in 2016.
林順喜教授為交大計算機工程系學士、臺大資工系碩士及博士,目前為臺灣師大資工系教授,研究興趣為演算法及人工智慧,是「外圍開局五子棋」及「師大大師中文輸入法」的發明人。截至2015年年底,林順喜教授共指導學生們參加國內外對局競賽共榮獲17面金牌。指導碩士畢業生共10位榮獲碩士論文獎。另總共指導5位博士畢業生,其中3位榮獲博士論文獎。其中最有名的是黃士傑(Aja Huang)博士,他是AlphaGo的主要研發者。AlphaGo於2016年打敗了人類棋王李世乭。
Last modification time:2016-03-29 AM 10:32

cron web_use_log