The game of Go has long been viewed as one of the most challenging tasks for artificial intelligence (AI) to master, so it is not surprising that last month's sensational victory by a Google AI named AlphaGo over world champion Lee Sedol drew global attention. In this talk, I use the 2016 Nature article on AlphaGo to facilitate a discussion on machine learning and AI. I cover their underlying concepts including, but not limited to, supervised learning, neural networks, Monte Carlo (MC) rollouts, and MC tree searches.

Reference:
[1] Silver D. et. al., Mastering the game of Go with deep neural networks and tree search, Nature 529, 484–489 (28 January 2016).

Prof. Jooyoung Lee is currently the Director of Center for In Silico Protein Science and Professor in School of Computational Sciences, in Korea Institute for Advanced Study. Prof. Lee got Ph.D. degree from Brown University in 1990, under Prof. John Michael Kosterlitz. Since 1990, Prof. Lee has published over 98 peer-reviewed papers in various SCI journals. Many of these papers are well cited [(over 3500 citations from SCI journals and over 3000 citations to the first/corresponding-authored papers (as of Mar. 2016)]. H-Index of Lee's publication is 29.