Re: [computer-go] New list member
Welcome on the list, Trevoke. :-) On Sun, Nov 01, 2009 at 08:20:15PM -0500, Aldric Giacomoni wrote: terry mcintyre wrote: In general, the time taken to run 1000-game tests hinders research. That's one avenue to explore better solutions, perhaps. Warm Regards, Terry McIntyre terrymcint...@yahoo.com Rene, Terry, Thanks for the warm welcome and the information! Terry, I'll have to explore that, but it seems to be a corollary of The Way Things Are more than anything else, isn't it? Well, it would be interesting to find a metric that is quicker to compute and shown to be strongly correlated to game tests in wide variety of conditions... It would be of immense practical value especially since it would enable much quicker and thus better tuning of various constants and such. However, the task itself seems quite hard to me and kind of janitorial in the field. Someone did send me a link to the Go Library - there is now tons of stuff to read in there. Most of the programs out there now are Neural Networks, it seems. Are there any who tried to play with knowledge hard-wired in there, such as Smart Go (http://www.smart-games.com/knowpap.txt) ? And if so, what is that knowledge? GNUGO is a prime example of popular program with hard-wired knowledge, though it's quite a few stones weaker than the strongest MCTS programs nowadays. I think more work on integrating MCTS in GNUGO in a smart way could make for a very interesting project. On the topic of academia.. While I would love to actually do graduate studies around the game of go, I may have to do something boring, like solving a real-world problem that would make people's lives better (gasp). Does anyone have an idea of real-world problems which could be correlated to go as far as AI goes (besides developing my own brew of Psychohistory) ? That depends on the particular method you choose to research, I guess. MCTS is generally usable for wide variety of planning tasks with difficult-to-create evaluation function and very wide search space, and MCTS Go research is often sold as a precursor of such. I'd be very interested in some concrete application examples as well, though. Does anybody here use Ruby at all for coding? Or is everyone using lower-level languages like C++ ? Java is somewhat popular, but most people use some variant of C I think, mainly since MCTS is performance-critical task. If you'd choose a different approach than MCTS, different language choices might make sense as well. -- Petr Pasky Baudis A lot of people have my books on their bookshelves. That's the problem, they need to read them. -- Don Knuth ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] New list member
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 terry mcintyre wrote: In general, the time taken to run 1000-game tests hinders research. That's one avenue to explore better solutions, perhaps. Warm Regards, Terry McIntyre terrymcint...@yahoo.com Rene, Terry, Thanks for the warm welcome and the information! Terry, I'll have to explore that, but it seems to be a corollary of The Way Things Are more than anything else, isn't it? Someone did send me a link to the Go Library - there is now tons of stuff to read in there. Most of the programs out there now are Neural Networks, it seems. Are there any who tried to play with knowledge hard-wired in there, such as Smart Go (http://www.smart-games.com/knowpap.txt) ? And if so, what is that knowledge? On the topic of academia.. While I would love to actually do graduate studies around the game of go, I may have to do something boring, like solving a real-world problem that would make people's lives better (gasp). Does anyone have an idea of real-world problems which could be correlated to go as far as AI goes (besides developing my own brew of Psychohistory) ? Does anybody here use Ruby at all for coding? Or is everyone using lower-level languages like C++ ? - -- Aldric Giacomoni Every civilization must contend with an unconscious force which can block, betray or countermand almost any conscious intention of the collectivity. -- Tleilaxu Theorem (unproven) -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkruM88ACgkQVcTYcihtgGJgyACg1yZ9AsdFmd9OoZVvXhy3rVYc 4NoAn0NjHmKzCHisoKxEZWPpFF5g4vxe =7LnW -END PGP SIGNATURE- ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
RE: [computer-go] New list member
Knowpap.txt is how The Many Faces of Go represents knowledge. Smart Go is a different program. Until a few years ago the strongest programs all used knowledge-intensive approaches with highly pruned local searches, like Many Faces. Now the strong programs all use Monte Carlo Tree Search, although Many Faces of Go is a hybrid. There are only a few programs that use Neural Networks, and they are not among the strongest. Regards, David Most of the programs out there now are Neural Networks, it seems. Are there any who tried to play with knowledge hard-wired in there, such as Smart Go (http://www.smart-games.com/knowpap.txt) ? And if so, what is that knowledge? ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
Re: [computer-go] New list member
Welcome Aldric, Not a frequent poster myself, here are two resources that you may find useful. 1) an extensive library of articles related to computer-go is collected on http://www.citeulike.org/group/5884/library This list provides a wealth of articles tracing back many years and used to be very current. More recently, not everybody is sending links of their papers, which is a pity as they are sometimes hard to find elsewhere. For Machine Learning papers in particular, Markus Enzenberger's Neurogo papers are highly recommended and a good starting point. But there are others as well, including several Ph.D. theses. Those are nice because they are (always) much longer than papers, (sometimes) more instructive, and contain long reference lists. 2) Erik van der Werf did excellent work for his thesis on Machine Learning algorithms for particular functions. His website is http://erikvanderwerf.tengen.nl/ and contains links to his papers. Good luck, René 2009/10/29 Aldric Giacomoni ald...@trevoke.net Hi everyone, I've been following the list for about a week and a half, and thought I ought to introduce myself. I don't know if this much activity is normal on the list, but I'm glad to see there is so much to read :) My name is Aldric - just in case you hadn't guessed. I am 27 years old, and I live on the east coast of the US. I can be found online ( godiscussions.com, KGS, Twitter, and many other non-go-related places) as Trevoke. I've been playing go since 2006, and have reached the rank of 6k on KGS so far. I graduated college in 2004 with a major in Computer Science and Mathematics. I've been studying isshinryu karate since 2004 and am preparing for my 2d test. I've always had an interest in Artificial Intelligence (I found recently some files on my machine dealing with machine learning and dating back to 2002), but never pushed it. I figured, oh, it's too complex, I'll study that someday. For various reasons, I've decided to do some graduate studies (a Doctorate) in AI. You should now be able to take a good guess at why I joined this list. I currently have zero knowledge of artificial intelligence, besides the few papers about MTCS, and the paper around Crazy Stone and such, by Remy Coulom, that I read in the past few days, following Olivier's message to this list. I'm waiting for a first order of books from Amazon.com to get my feet wet.. I know one thing, which I am also aware is still vague: I want to help solve go. I realize this may involve some pattern recognition, knowledge representation, and a few more topics.. But I know this is where I want to go with it. Are there any resources which you could recommend to someone who would like to learn? Any pitfalls you can recommend I avoid (or stumble into) ? As a warning, I am an avid reader and a rather obstinate individual when I have decided to study/learn something. Oh, and finally, besides Olivier's suggestion to apply for a Doctorate with his team, do you know of anybody in the world who may consider taking someone like me (fluent in French, English, Italian, no ties, can travel, etc etc) in their team, to do that kind of work ? And.. If so, the repetition of the earlier question: what do you think I need to know before I can study with them? Thanks in advance :-) --Aldric ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/ ___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
[computer-go] New list member
Hi everyone, I've been following the list for about a week and a half, and thought I ought to introduce myself. I don't know if this much activity is normal on the list, but I'm glad to see there is so much to read :) My name is Aldric - just in case you hadn't guessed. I am 27 years old, and I live on the east coast of the US. I can be found online (godiscussions.com, KGS, Twitter, and many other non-go-related places) as Trevoke. I've been playing go since 2006, and have reached the rank of 6k on KGS so far. I graduated college in 2004 with a major in Computer Science and Mathematics. I've been studying isshinryu karate since 2004 and am preparing for my 2d test. I've always had an interest in Artificial Intelligence (I found recently some files on my machine dealing with machine learning and dating back to 2002), but never pushed it. I figured, oh, it's too complex, I'll study that someday. For various reasons, I've decided to do some graduate studies (a Doctorate) in AI. You should now be able to take a good guess at why I joined this list. I currently have zero knowledge of artificial intelligence, besides the few papers about MTCS, and the paper around Crazy Stone and such, by Remy Coulom, that I read in the past few days, following Olivier's message to this list. I'm waiting for a first order of books from Amazon.com to get my feet wet.. I know one thing, which I am also aware is still vague: I want to help solve go. I realize this may involve some pattern recognition, knowledge representation, and a few more topics.. But I know this is where I want to go with it. Are there any resources which you could recommend to someone who would like to learn? Any pitfalls you can recommend I avoid (or stumble into) ? As a warning, I am an avid reader and a rather obstinate individual when I have decided to study/learn something. Oh, and finally, besides Olivier's suggestion to apply for a Doctorate with his team, do you know of anybody in the world who may consider taking someone like me (fluent in French, English, Italian, no ties, can travel, etc etc) in their team, to do that kind of work ? And.. If so, the repetition of the earlier question: what do you think I need to know before I can study with them? Thanks in advance :-) --Aldric___ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/