MATLAB interface for SVM-Light

beta
3.0 Stars (2)
20 Downloads (This Week)
Last Update:
Download Matlab R14 binary release
Browse All Files

Description

mex-svm is a set of patches against SVM-Light to compile into "mex" libraries and enable fast Support Vector Machine evaluation from within MATLAB.

MATLAB interface for SVM-Light Web Site

Update Notifications





User Ratings

★★★★★
★★★★
★★★
★★
1
0
0
0
1
ease 1 of 5 2 of 5 3 of 5 4 of 5 5 of 5 3 / 5
features 1 of 5 2 of 5 3 of 5 4 of 5 5 of 5 5 / 5
design 1 of 5 2 of 5 3 of 5 4 of 5 5 of 5 4 / 5
support 1 of 5 2 of 5 3 of 5 4 of 5 5 of 5 3 / 5
Write a Review

User Reviews

  • nhammes
    1 of 5 2 of 5 3 of 5 4 of 5 5 of 5

    Works as described after running the provided patch. The first couple of times I attempted to use it, MATLAB closed unexpectedly without error. After these attempts, it has worked without issue since. Thanks for the well-written interface.

    Posted 07/29/2014
  • oid-3938287
    1 of 5 2 of 5 3 of 5 4 of 5 5 of 5

    The author correctly recognizes that Matlab has a native Compressed Sparse Column (CSC) format for sparse matrices. He assumes that users prefer to store data in row format, so each pattern is a row of the data matrix; but this matrix is stored as a CSC matrix. The underlying data structure for SVM Light uses a sparse data structure for each pattern. To convert between representations, a matrix transpose is needed. This author uses the worst possible algorithm for this (in mexcommon.c)---it takes m x n time! There are much faster and better ways to do this, like just using Matlab's built-in, highly-optimized transpose method. If you have dense data, this code will work fine for you. If you have sparse data, you'll be mostly out of luck. There's a lot of good code here, but I had to completely rewrite mexcommon.c.

    Posted 08/07/2012
Read more reviews

Additional Project Details

Intended Audience

Science/Research

Programming Language

C, MATLAB

Registered

2008-01-22
Screenshots can attract more users to your project.
Features can attract more users to your project.

Icons must be PNG, GIF, or JPEG and less than 1 MiB in size. They will be displayed as 48x48 images.