R: Is there an alternative SVM implementation than libsvm in e1071 package?

libsvm, r, svm

Solution

Check the kernlab package:

kernlab: Kernel-based Machine Learning Lab

Kernel-based machine learning methods for classification, regression, clustering, novelty detection, quantile regression and dimensionality reduction. Among other methods kernlab includes Support Vector Machines, Spectral Clustering, Kernel PCA, Gaussian Processes and a QP solver.

Kernlabs `ksvm` supports C-svc, nu-svc, (classification) one-class-svc (novelty) eps-svr, nu-svr (regression) formulations along with native multi-class classification formulations and the bound-constraint SVM formulations. `ksvm` also supports class-probabilities output and confidence intervals for regression.

An interface to the `SVMlight` implementation is provided in package klaR

See also the CRAN Task View Machine Learning & Statistical Learning

Problem

I try to compare different implementations of SVMs in R. Is there another one than the libsvm implementation in the e1071 package ? Generally, is there a good alternative from libsvm which implements the nu-SVM and epsilon-SVM ?

Original source