It is currently Fri Mar 29, 2024 1:26 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 3 posts ] 
Author Message
Offline
 Post subject: New paper: Multi-Labelled Value Networks for Computer Go
Post #1 Posted: Wed Jun 07, 2017 2:26 pm 
Gosei

Posts: 1590
Liked others: 886
Was liked: 527
Rank: AGA 3k Fox 3d
GD Posts: 61
KGS: dfan
I just opened this up but it looks pretty interesting. https://arxiv.org/abs/1705.10701

Abstract:
Quote:
This paper proposes a new approach to a novel value network architecture for the game Go, called a multi-labelled (ML) value network. In the ML value network, different values (win rates) are trained simultaneously for different settings of komi, a compensation given to balance the initiative of playing first. The ML value network has three advantages, (a) it outputs values for different komi, (b) it supports dynamic komi, and (c) it lowers the mean squared error (MSE). This paper also proposes a new dynamic komi method to improve game-playing strength. This paper also performs experiments to demonstrate the merits of the architecture. First, the MSE of the ML value network is generally lower than the value network alone. Second, the program based on the ML value network wins by a rate of 67.6% against the program based on the value network alone. Third, the program with the proposed dynamic komi method significantly improves the playing strength over the baseline that does not use dynamic komi, especially for handicap games. To our knowledge, up to date, no handicap games have been played openly by programs using value networks. This paper provides these programs with a useful approach to playing handicap games.


This post by dfan was liked by: Kirby
Top
 Profile  
 
Offline
 Post subject: Re: New paper: Multi-Labelled Value Networks for Computer Go
Post #2 Posted: Wed Jun 07, 2017 3:02 pm 
Honinbo

Posts: 9545
Liked others: 1600
Was liked: 1711
KGS: Kirby
Tygem: 커비라고해
Seems like their main idea, in contrast to AlphaGo, is to have the value network output a vector of win rates for different komi values, rather than just the single win rate that AlphaGo computes.

_________________
be immersed

Top
 Profile  
 
Offline
 Post subject: Re: New paper: Multi-Labelled Value Networks for Computer Go
Post #3 Posted: Wed Jun 07, 2017 6:45 pm 
Lives in sente

Posts: 727
Liked others: 44
Was liked: 218
GD Posts: 10
DeepZenGo will probably need this paper the most, as it has to play in MLily Cup this month. Though MLily Cup uses Chinese rule which most programs don't have any problems with. And it's already much stronger than CGI Go intelligent. Latest Zen running on just single core of CPU achieve higher Elo than CGI running on quad-core.

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 3 posts ] 

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group