Yuhao Yang May 2016

SVMOnSpark is an implementation of distributed SMO algorithm to train a binary Support Vector Machine. The scalability is supposed to be very good as it avoids shuffle and unnecessary communication.

It also supports arbitrary kernels. Currently linear and RBF are embedded.

Typical Spark ml pattern. SVM is an Estimator and SVMModel is the corresponding model. You can also refer to the Example folder.

The implementation has been tested against MNist dataset and get an accuracy about 99% on the test data set (one vs rest). We'll post more results once they are ready.