slack variables from nu-svm

[Eugen Funk]

Dear mlpy team,

in my project I would like to access the slack variables (v) from the nu-svm formulation:

min  w^Tw+C*\sumv
s.t. y_i(w^Tx_i+b)>=1-v
v_i>0

Is there a way to do so?
Until now I reimplemented the optimization using cvxopt, however the solution is quite slow and it takes a lot of time if I try to run many experiments.


Thank you in advance for the hints!