The free tools listed below range from C++ libraries, to drag and drop graphical environments. Some are pure support vector machine (SVM) and others are data mining platforms which include SVM.
KNIME – broad functionality GUI based data mining tool with particularly easy to use SVM support.
LIBSVM – is a sophisticated SVM library containing C-classification, v-classification, epsilon-regression, and v-regression. It supports multi-class classification, weighted SVM for unbalanced data, cross validation and automatic model selection. Interfaces for Python, R, S+, MATLAB, Perl, Ruby and LABVIEW make it particularly popular.
mySVM – written in C++ and includes SVM classification and regression. Avaliable as source code and WIndows binaries. Kernels include linear, polynomial, radial basis function, neural(tanh) and anova.
Orange – is another GUI based data mining toolset that includes a SVM learner.
RapidMiner – is a sophisticated GUI based data mining toolset, and possibly the most widely used platform of its type. It includes and includes several SVM implementations (including LIBSVM).
SVM Light – is a very widely used classification and regression package and is distributed as C++ source. It is well known for its speed of execution and an efficient implementation of the leave-one-out (LOO) cross-validation method.
Weka – one of the best known collection of data mining tools contains an SVM implementation.