Project details for SGD

Logo SGD 1.2

by leonbottou - September 24, 2008, 17:09:35 CET [ Project Homepage BibTeX Download ]

1248 views (5 today), 282 downloads ( 1 today ), 1 comment, 1 subscription

OverallWhole StarWhole StarWhole Star1/2 StarEmpty Star
FeaturesWhole StarWhole StarEmpty StarEmpty StarEmpty Star
UsabilityWhole StarWhole StarWhole StarWhole StarEmpty Star
DocumentationWhole StarWhole StarWhole StarWhole StarEmpty Star
(based on 1 vote)
Description:

The SGD package contains a stochastic gradient implementation of linear SVMs and linear CRFs. It demonstrate that a simple stochastic gradient descent is very competitive algorithm for such tasks. For instance, the SVMSGD program digests the RCV1 database in slightly more than one second, and the CRFSGD code runs close to ten times faster then the CRF++ Limited Storage BFGS optimizer.

BibTeX Entry:
Download
URL:
Project Homepage
Supported Operating Systems:
Cygwin, Linux, Macosx, Windows
Tags:
Large Scale, Svm, Crf, Stochastic Gradient Descent
Archive:
download here

Comments

Leon Bottou (on September 23, 2008, 21:42:34)

Version 1.2 fixes bug in the preprocessing program.

Leave a comment

(will not appear publicly)

You may use Markdown syntax here, but raw HTML will be removed.