Passive-Agressive-SVM-for-online-learning

An implementation of Passive-Agressive version of SVM algorithm used for online learning.

The implementation follows Scikit-Learn API.

This implementation allows to use Kernelized Version of SVM and different optimiation strategies.

Avaliable Kernel Options: linear, rbf, polynomial Avaliable Optimization Strategies:

• tau = loss / ||X||^2
• tau = min(c,loss / ||X||^2)
• tau = loss/(||X||^2 + (1/(2*C)))