Closed MNLubov closed 1 year ago
Hi @MNLubov,
the model does indeed save the features only for the support vectors. If you increase the cost
parameter (to e.g. 100.
), you should see that:
a) model.SVs.l
becomes smaller/the second dimension of model.SVs.X
becomes smaller.
b) the serialized model file becomes smaller.
Hope this helps!
Hi @till-m
Thanks for the clarification!
I am training SVM on quite big dataset, and after training I save the model. The size of the model depends on the size of the dataset. The size of my training data is 12MB and size of the model is 66 MB. Do I understand correctly that this depends on number of the support vectors and that model doesn't save training data itself?