a*****g 发帖数: 19398 | 1 A computer program that can defeat a professional human player at the
classic strategy game Go is reported in this week’s Nature. Go has been
regarded as an outstanding ‘grand challenge’ for artificial intelligence
owing to its large search space and the difficulty of evaluating board
positions and moves. The findings provide hope that human-level performance
could potentially be achieved in other seemingly intractable artificial
intelligence domains.
In the game Go, which originated in ancient China, two players alternately
place black and white pieces onto a square grid with the aim of occupying
more territory than their opponent at the end of the game. The most
successful computer Go programs developed thus far play at the level of
human amateurs and have not been able to defeat a human professional player
in even games.
David Silver, Aja Huang and Demis Hassabis and colleagues developed a
program called AlphaGo that uses ‘value networks’ to evaluate board
positions and ‘policy networks’ to select moves. These deep neural
networks are trained through a combination of supervised learning from human
expert games and reinforcement learning from games it plays against itself.
AlphaGo achieved a 99.8% winning rate against other Go programs and
defeated the human European Go champion in a tournament by 5 games to 0.
This is the first time a computer program has defeated a professional player
in the full-sized game of Go with no handicap — a feat previously thought
to be a decade away.
AlphaGo’s next challenge will be to play Mr Lee Sedol — who is
acknowledged as the top Go player in the world over the past decade — in
Seoul in March. | d****g 发帖数: 7460 | | b***d 发帖数: 2695 | 3 当有一天人类被人类自己发明创造的机器人defeat的时候,
是不是很作的一件事啊?
:)
performance
【在 a*****g 的大作中提到】 : A computer program that can defeat a professional human player at the : classic strategy game Go is reported in this week’s Nature. Go has been : regarded as an outstanding ‘grand challenge’ for artificial intelligence : owing to its large search space and the difficulty of evaluating board : positions and moves. The findings provide hope that human-level performance : could potentially be achieved in other seemingly intractable artificial : intelligence domains. : In the game Go, which originated in ancient China, two players alternately : place black and white pieces onto a square grid with the aim of occupying : more territory than their opponent at the end of the game. The most
|
|