User:Albert/Archive Discussion

From Robowiki
< User:Albert
Revision as of 10:50, 1 April 2009 by Nat (talk | contribs) (talk archive)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
       Archive        This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.     

Albert. I just want to say that I am amaxed by the cheer volume of knowledge and thoughts around Robocode you share. Your latest initiative with a BeginnersFAQ is great. I'll try to free some time to try to add to it as well, though I might be short of time this weekend. And this FAQ is just the last of many, many pages and posts you contribute with. The official Alphaworks forum wouldn't be the same without you ready to answer questions ranging from newbie-type to advanced concepts. And this wiki is as useful as it is much thanks to you. I could never have wished for better help in coming up to speed in Robocoding than what I have go from your wiki-contributions. -- PEZ

Well... I had some spare time lately, and also my *fast* computer crashed and it will take me some time to get a new one, so in the mean time I navigate the forums. Also, I think the Wiki is the greatest Robocode resource out there, because it is the first centralized/permanent/easily maintenable Robocode knowledge site. So I'm happy to contribute. -- Albert

Hey, question for you, Albert - You described LauLectrik as having a "Best P-Space" gun - does that mean it's more of a statistical approach than a pattern approach? -- Kawigi

True. It is a pure statistical approach (you segment the data, build the probability functions) and fire to the most probable bearing. The only difference between BPS and classical statistical approach is that BPS saves many families of probability functions (pe. by distance and heading, by distance and accumulated velocity, etc) and always uses the best for a given moment. But it nevew worked as expected... because the ammout of data you need to make all this information is so high that learns very slowly. -- Albert

About this probability functions. I'm trying to include some bayes nets in robocode. The problem is that i'm new in this and for now everithing is floating in a dense fog for me. You sed you don't want to share the source but you are willing to give some answers. Can you help me on building this probabilities? Can i contact you by email or how? --H7

Albert, is there any way that I can get into contact with you via email? I am seriously considering updating RoboRumble@Home so it will be able to run with current Robocode versions. In fact, I already did an alpha version that runs with Robocode 1.2.5A, but has some issues. Is it okay with you that I update RoboRumble@Home? Should we create a project page on SourceForge like it's done with Robocode? Do you have any objections, restrictions, ideas or similar? :-)  --Fnl