Hyperparameters exposure

[gcattan]

Ok. Like SVC, QuanticSVM should allow to set SVM hyper-parameters in its init.

Originally posted by @qbarthelemy in #2 (comment)

[gcattan]

Todo list:

• gamma (SklearnSVM)
• feature_map (QSVM and VQC)
• number of shots (QSVM and VQC)
• optimizer (VQC)
• variational form (VQC)
• enforce_spd (QSVM and VQC)
• number of iteration (QSVM, VQC, SklearnSVM)
• output normalization (QSVM, VQC, SklearnSVM)
• L2 regularisation (QSVM, VQC, SklearnSVM)

[qbarthelemy]

Can we close this issue?

More generally, do not forget to close issues when PR are merged.

[gcattan]

Hello,

I was not sure how to manage stories.
There are four parameters to expose again, but I can create child issues and reference them in this one if you prefer.

[gcattan]

After double-checking, it seems that these last parameters are not accessible directly in the version of qiskit we used.
I think we can safely close this issue and add follow-up for C and max_iter parameters, after #15.

[qbarthelemy]

Ok.