Advanced Search Tips


Submitted by on Jul 31 2014 Suggest Revision
By: Chih-Chung Chang and Chih-Jen Lin
From: National Taiwan University
Resource Type:
LIBSVM (modified BSD)
C++, Java
Data Format:


LIBSVM is a popular open source machine learning libraries, developed at the National Taiwan University and written in C++ though with a C API. LIBSVM implements the SMO algorithm for kernelized support vector machines (SVMs), supporting classification and regression.
Categorized in: Machine Learning
Post comment
External User
2014-09-14 04:31:45

It is well written and the de facto implementation of svms George Tzanetakis

External User
2014-09-12 14:45:34

libSVM is open source, and it is fast, reliable, and provides lots of SVM functionalities. Nevertheless, it is not thread safe, which may render it useless for some applications. Although I contacted the authors with some fixes that would solve it, they were not interested. Lluís Padró

External User
2014-09-12 07:36:10

Efficient and complete implementation of many SVM-based algorithms. G. Loosli

External User
2014-09-11 19:17:09

Reliable, robust implementation of nonlinear SVMs. Use Liblinear instead for linear models. Be sure to read the practical guide from the authors. Charles Elkan

External User
2014-09-11 17:44:27

motaz el Saban

External User
2014-09-11 17:36:15

Easy to use, well documented Jennifer Tam

External User
2014-09-10 23:29:51

Easy to use and scalable. Cho-Jui Hsieh