Doubt in Grid Search

#1

I am getting the following output while computing the best estimator in grid Search -

SVC(C=5.0, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,
decision_function_shape=‘ovr’, degree=3, gamma=‘scale’, kernel=‘rbf’,
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)

Is it fine as it is completely opposite to what is shown in the video?
Also, the accuracy I got using cross_val_score(svc,X,Y,scoring = ‘accuracy’,cv = 5).mean()
was 0.96 while the in the video it was very low around 42 %.

#2

hey @sanchit123manchanda ,
it just depends upon what parameters you provide for grid search. Based in those different sets of parametes,grid search will find the best combination of parameters which will score the highest among them.
So now , you found such parameters which are able to achieve cross validation score as 0.96 and that is really really good. Very nice.

In video also , 42-44 % score was achieved with defult parameters in SVC and after tuning we were able to achieve around 97% . So just the important thing in grid search is what range of parameters you provide to find the best combination.

its fine absolutely.

I hope this would have resolved your doubt.
Thank You and Happy Learning :slightly_smiling_face:.