Comments on: Symbols as Hash Codes for Super-fast Linear Classifiers http://lingpipe-blog.com/2008/03/25/symbols-as-hash-codes-for-super-fast-linear-classifiers/ Alias-i’s Natural Language Processing and Text Analytics API Wed, 20 Aug 2008 23:59:15 +0000 http://wordpress.org/?v=MU By: An Online Logistic Regression API with Regularization « LingPipe Blog http://lingpipe-blog.com/2008/03/25/symbols-as-hash-codes-for-super-fast-linear-classifiers/#comment-2457 An Online Logistic Regression API with Regularization « LingPipe Blog Tue, 03 Jun 2008 23:13:20 +0000 http://lingpipe.wordpress.com/?p=83#comment-2457 [...] so a constant stream of new features won’t blow out memory (see my previous blog entry on using hash codes as features). Unfortunately, VW runs from the command-line and requires the training examples to be in a data [...] [...] so a constant stream of new features won’t blow out memory (see my previous blog entry on using hash codes as features). Unfortunately, VW runs from the command-line and requires the training examples to be in a data [...]

]]> By: Feature Hash Code Collisions in Linear Classifiers « LingPipe Blog http://lingpipe-blog.com/2008/03/25/symbols-as-hash-codes-for-super-fast-linear-classifiers/#comment-2164 Feature Hash Code Collisions in Linear Classifiers « LingPipe Blog Fri, 02 May 2008 18:38:56 +0000 http://lingpipe.wordpress.com/?p=83#comment-2164 [...] a comment to John Langford’s response to my blog entry recapitulating his comments after his talk, Kuzman Ganchev points out that he and Mark Dredze did [...] [...] a comment to John Langford’s response to my blog entry recapitulating his comments after his talk, Kuzman Ganchev points out that he and Mark Dredze did [...]

]]> By: Machine Learning (Theory) » Eliminating the Birthday Paradox for Universal Features http://lingpipe-blog.com/2008/03/25/symbols-as-hash-codes-for-super-fast-linear-classifiers/#comment-2138 Machine Learning (Theory) » Eliminating the Birthday Paradox for Universal Features Sat, 26 Apr 2008 18:40:49 +0000 http://lingpipe.wordpress.com/?p=83#comment-2138 [...] want to expand on this post which describes one of the core tricks for making Vowpal Wabbit fast and easy to use when learning [...] [...] want to expand on this post which describes one of the core tricks for making Vowpal Wabbit fast and easy to use when learning [...]

]]> By: craig http://lingpipe-blog.com/2008/03/25/symbols-as-hash-codes-for-super-fast-linear-classifiers/#comment-2076 craig Fri, 04 Apr 2008 19:38:42 +0000 http://lingpipe.wordpress.com/?p=83#comment-2076 http://weblogs.java.net/blog/kohsuke/archive/2008/03/deep_dive_into.html This article gives a method for getting the assembly code of native code produced by the JIT. Take a peek. http://weblogs.java.net/blog/kohsuke/archive/2008/03/deep_dive_into.html

This article gives a method for getting the assembly code of native code produced by the JIT. Take a peek.

]]>