Re: [Computer-go] Facebook Go AI
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Am 06.12.2015 um 16:24 schrieb Petr Baudis: > On Sat, Dec 05, 2015 at 02:47:50PM +0100, Detlef Schmicker wrote: >> I understand the idea, that long term prediction might lead to a >> different optimum (but it should not lead to one with a higher >> one step prediction rate: it might result in a stronger player >> with the same prediction rate...) > > I think the whole idea is that it should improve raw prediction > rate on unseen samples too. This would mean, we are overfitting, which should be seen by a bigger difference between unseen and seen samples, which is normaly checked by using a test database and compare the results with the train database and small ?! The motivation of the increased suprevision is > to improve the hidden representation in the network, making it > more suitable for longer-term tactical predictions and therefore > "stronger" and better encompassing the board situation. This > should result in better one-move predictions in a situation where > the followup is also important. > > It sounds rather reasonable to me...? > Yes, it sounds reasonable, but this not always helps in computer go :) Detlef -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.22 (GNU/Linux) iQIcBAEBAgAGBQJWZGoGAAoJEInWdHg+Znf40Z0P/3vHdeHIb9/5Fqp78hdT45IC RizA7703fWMeU8JC6dZA1ziI/oKfGLTetFVFGIcPIx+T3lkRapZLYZNa8BXQGXZr lnjSk2aEfsJdCZd+Y62ECLitXEOOosjvF9bHNoAj39MeuDOUMxBcjCSSUjNDbOTm eYWAYC0UgTE7xzR629FHQ1PJ6+iw2RYsvfEmEXwCD2blT8gab4uNDZelk+R/l4KI F73Sid8ULz3AE8zi6XX2qWHmupKa7bMI9ZWcDfAf78rSPMOx6SPYbrX/QXsCiQAD fy3KNOro93HJyW1sDk4mJEERm+UjmOcGjxILQDkcRo/+D/SdLT2DbeXM1KuE3RWu 0tSsubBNvNv5begSLkhOvSrybKZYtsgTqycF+4dMzLVj3LO5+Iy957w1QfwHBjBZ ATT4GFPidrnrGfdKcRL/mtQJi0+JZ/QVZoIMvxYHMVk6Vz9JRwGFdFZKL1mnX5oy t9Y8tIzkkQnjJk4/XRVuDnngBKFFgBgbyLWy6WU1MmhEGNyHt7anUKDyxHHz4LU/ 00vJUIGiMsaOTqT92yLPWOm39RHRP2J9Pfflz12+ysGpD+VKsSNZOm6wdyn8/q5c A3zrVJM7tug2Iu7mjJ6BC10bF813EGKfgtjxrQ54TnlIucMDssv9rBajqsoRDIxG 14Iv+3P/YDY0hmRoI51S =Bx1M -END PGP SIGNATURE- ___ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
Re: [Computer-go] Facebook Go AI
On Sat, Dec 05, 2015 at 02:47:50PM +0100, Detlef Schmicker wrote: > I understand the idea, that long term prediction might lead to a > different optimum (but it should not lead to one with a higher one > step prediction rate: it might result in a stronger player with the > same prediction rate...) I think the whole idea is that it should improve raw prediction rate on unseen samples too. The motivation of the increased suprevision is to improve the hidden representation in the network, making it more suitable for longer-term tactical predictions and therefore "stronger" and better encompassing the board situation. This should result in better one-move predictions in a situation where the followup is also important. It sounds rather reasonable to me...? -- Petr Baudis If you have good ideas, good data and fast computers, you can do almost anything. -- Geoffrey Hinton ___ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
Re: [Computer-go] Facebook Go AI
> I understand the idea, that long term prediction might lead to a > different optimum (but it should not lead to one with a higher one > step prediction rate: it might result in a stronger player with the > same prediction rate...), and might increase training speed, but hard > facts would be great before spending a GPU month into this :) > > > > Detlef > I wouldn't be too sure, that this cannot improve the 1 step prediction rate. In a way, multi-step prediction is like peeking around the corner. :-) Stefan ___ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
[Computer-go] Introductory computer go/MCTS presentation: slides and video
Hi there lovely computer-go mailing list, I gave a presentation titled "Beating Go Thanks to the Power of Randomness" at Rubyconf 2015. It is a full introductory talk which means: * the rules of Go are introduced (leaving out Ko and seki though.. time) * the discussion of MCTS is on a very high level, RAVE/AMAF is mentioned and heuristics are only briefly discussed * A lot of it focusses around the complexity of Go, why can't we beat it yet and some Chess vs. Go is included So it's not useful for anyone that has been on here for some time. If you want to share some introductory material with friends or so or for total newcomers it might be useful. Slides + video: https://pragtob.wordpress.com/2015/11/21/slides-beating-go-thanks-to-the-power-of-randomness-rubyconf-2015/ Also this was before we got further information about the Facebook Go AI, so that information is less than complete and I didn't yet know how strong it was (wow 3d KGS! :) ) Otherwise the information is up to date and correct to the best of my knowledge, sorry for any mispronunciations. If a more experienced Go programmer had the time to watch it and give feedback on the talk of any kind, I'd be very happy about that! Tobi -- www.pragtob.info signature.asc Description: OpenPGP digital signature ___ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go
