-
conner_bw committed revision 620 to the sux0r SVN repository, changing 5 files.
2009-10-30 01:26:14 UTC in sux0r
-
conner_bw changed the public information on the XRNS-PHP project.
2009-10-11 19:48:37 UTC in XRNS-PHP
-
conner_bw changed the public information on the XRNS-PHP project.
2009-10-11 19:48:35 UTC in XRNS-PHP
-
conner_bw committed revision 619 to the sux0r SVN repository, changing 1 files.
2009-10-02 02:34:38 UTC in sux0r
-
conner_bw committed revision 618 to the sux0r SVN repository, changing 21 files.
2009-10-01 14:12:01 UTC in sux0r
-
First of all, I believe vectors are skewed. If you have an uneven distribution of documents (50 in one cateorgy, 5 in another) then the data will be less accurate.
As for the percentage, please see the ./includes/suxNaiveBayesian.php file.
I am not a mathematician. My implementation is based on examples by Loic d'Anterroches and Ken Williams. URLS to their respective project pages are at...
2009-09-14 15:43:56 UTC in sux0r
-
Hi Steve,
Glad to hear you are using Sux0r.
1) Please see the ./includes/suxNaiveBayesian.php file.
I am not a mathematician. My implementation is based on examples by Loic d'Anterroches and Ken Williams. URLS to their respective project pages are at the top of that file. You'll will want to look at the categorize() and rescale() methods.
You are misinterpreting logspace scaling...
2009-09-11 17:25:54 UTC in sux0r
-
conner_bw committed revision 617 to the sux0r SVN repository, changing 1 files.
2009-09-10 19:02:36 UTC in sux0r
-
I think, perhaps, you are confusing the sux0r project with the goals of the ICBL Bayesian Feed Filtering project? Maybe you are contacting the wrong people?
The URL you provided is an RSS feed. Yes it will work fine. Other than the specifics of the content, there's no difference between the URL you provided or any other RSS feed.
Best regards.
2009-09-10 12:59:19 UTC in sux0r
-
conner_bw committed revision 616 to the sux0r SVN repository, changing 1 files.
2009-09-01 19:44:20 UTC in sux0r
