Google AlphaGo beats professional go player board game




Google has developed a computer program that is able to beat a professional player in the game of go. The program AlphaGo defeated the European champion of the game, Fan Hui, five out of five times.

Go is seen as one of the most challenging games for classic artificial intelligence to learn, due to the large search space, and the difficulty of determining positions and plate set. The software was developed by Deep Mind, a company that Google bought in 2014. How AlphaGo was able to win in the complex game set put in a study in Nature.

AlphaGo has two networks, one of which predicts the next move of the opponent, while the other is the result of several on the board attempts to predict. These networks are combined with an artificial intelligence algorithm to look ahead in the game looking for possible moves.

AlphaGo used value networks to choose to take on board the positions and policy networks for its next turn. The program features neural networks that are trained by learning games that have been played by experts and learning the games played by the program itself. The policy network suggests the best moves to make, while the value networks evaluate the extracted positions. AlphaGo then chooses the move that is most successful in his simulations. The self-played games stem from a search algorithm that uses the Monte Carlo method, which the program simulates thousands of games.

Go is a complex game in which two players each try to surround territory and try to win. The game has a player at any time put a choice of 200, set in comparison to 20 in chess. As a result, according to the Google researcher David Silver is not possible to all options with brute force surveys. Instead AlphaGo makes use of a more human approach, where the possibilities are limited. This makes the place seem more based on a system similar to human intuition.

The combination of these ensured that AlphaGo was able to get the go-European champion Fan Hui beat five times in a row. The program also has a 99.8 percent success rate against other go-programs. In March AlphaGo going up against one of the world champions of Go, Lee Sedol.


In: Technology & Gadgets Asked By: [15207 Red Star Level]

Answer this Question

You must be Logged In to post an Answer.

Not a member yet? Sign Up Now »