{"id":5224,"date":"2016-01-28T03:02:07","date_gmt":"2016-01-28T03:02:07","guid":{"rendered":"http:\/\/www.kurzweilai.net\/?p=272197"},"modified":"2016-01-29T09:06:34","modified_gmt":"2016-01-29T09:06:34","slug":"google-machine-learning-system-is-first-to-defeat-professional-go-player","status":"publish","type":"post","link":"https:\/\/hoo.central12.com\/fugic\/2016\/01\/28\/google-machine-learning-system-is-first-to-defeat-professional-go-player\/","title":{"rendered":"Google machine-learning system is first to defeat professional Go player"},"content":{"rendered":"
\"\"

Go is played on a grid of black lines (usually 19\u00d719). Game pieces, called stones, are played on the line intersections. (credit: Goban1\/Wikipedia)<\/p><\/div>\n

A deep-learning computer system called AlphaGo created by Google’s DeepMind team has defeated reigning three-time European Go champion Fan Hui 5 games to 0 — the first time a computer program has ever beaten a professional Go player, reports Google Research blog<\/a> today (Jan. 27) — a feat previously thought to be at least a decade away.<\/p>\n

“AlphaGo uses general machine-learning techniques to allow it to improve itself, just by watching and playing games,” according to\u00a0David Silver and Demis Hassabis of Google DeepMind. Using a vast collection of more than 30 million Go moves from expert players, DeepMind researchers trained their system to play Go on its own.<\/p>\n

To achieve that, AlphaGo, as described in a paper in\u00a0Nature<\/a> today, combines a state-of-the-art tree search with two\u00a0deep neural networks<\/a>, each containsing many layers with millions of neuron-like connections needed to deal with Go’s vast search space — more than a\u00a0googol<\/a> (10100<\/sup>) times larger than chess (a number greater than there are atoms in the universe).<\/p>\n

“We first trained [one of the two networks] on 30 million moves from games played by human experts, until it could predict the human move 57% of the time …. But our goal is to beat the best human players, not just mimic them, Silver and Hassabis said. “To do this, AlphaGo learned to discover new strategies for itself, by playing thousands of games between its neural networks, and gradually improving them using a trial-and-error process known as\u00a0reinforcement learning<\/a>.”<\/p>\n

\"\"<\/a>

This figure from the Nature article shows the Elo rating (a 230 point gap corresponds to a 79% probability of winning) and approximate rank of AlphaGo (both single machine and distributed versions), the European champion Fan Hui (a professional 2-dan), and the strongest other Go programs, evaluated over thousands of games. Pale pink bars show the performance of other programs when given a four move headstart. (credit: David Silver et al.\/Nature)<\/p><\/div>\n

AlphaGo\u2019s next challenge will be to play the top Go player in the world over the last decade,\u00a0Lee Sedol<\/a>. The match will take place this March in Seoul, South Korea.<\/p>\n

“While games are the perfect platform for developing and testing AI algorithms quickly and efficiently, ultimately we want to apply these techniques to important real-world problems,” the researchers say. “Because the methods we have used are general purpose, our hope is that one day they could be extended to help us address some of society\u2019s toughest and most pressing problems, from climate modelling to complex disease analysis.”<\/p>\n