Add Support Vector Machines to Weka, and experiment with it.

[damirsaleschoice]

Recall the Lotto 6/49 ticket, as well as the Neural Networks ticket. You will apply that knowledge here...

1. Add the SVM to Weka (that is LibSVM)
2. Experiment with different SVM types (properties)
3. Find which types are good with regression, and which types are good for classification.

We will use this for the ML component of our final application.

[sherlockliang888]

RBF is one of the most commonly used kernel functions in svm. When there is no prior knowledge of data and data is non-linear, we could start with this. Linear kernel is most efficient for text classification. Sigmoid kernel would be similar to NN, since it's using tanh activation function. These three kernels can be experimented around for sentiment analysis problem.

Done

Need PR

[damirsaleschoice]

Good work!