SVM and Parameter Optimization with GridSearchCV
ฝัง
- เผยแพร่เมื่อ 11 เม.ย. 2019
- Note that data in the Cancer Research file has similarly scaled attributes due to the measurement systems. Hence, I did not run a scaler to transform the data. SVM might be impacted by scale differences in attributes. Hence, it is always a good idea to check data to see if you need to transform the data and scale it before you run SVM.
During the GridSearchCV when I defined " C':[1,42,10]" (around 24th-minute mark) that line inside the Param only checks C values of 1, 42, and 10 rather than my intended purpose of checking values between 1 to 41 with an increment of 10. The correct way is to define a Numpy array as given below for C and coef0.
param={'kernel':('linear', 'poly', 'rbf', 'sigmoid'),
'C':np.arange(1,42,10),
'degree':np.arange(3,6),
'coef0':np.arange(0.001,3,0.5),
'gamma': ('auto', 'scale')}
it will be great that you can provide more videos like this.... thanks for share this one
Thanks for the nice explanation. Crystal clear!!
Very well explained. Thank you so much.
Excellent!
Thank you
I am getting error at 7th line i.e valueError: unknown label "continuous"
Hello, please do you also have a tutorial for SVR parameter optimization?..i'm having a hard time with SVR
I got not fitted error while running confusion_matrix
Thanks for this video. I tried these two ways with my data, but python has been running for more than 60 hours and still, I couldn't get a result. Is this possible, or is there any mistake?
It really depends on the data size and your computer speed. You can try the algorithm with a small portion of the data to make sure your code works. Then you can upgrade to a faster computer or just wait.
@@ismailcapar1103 sir, please provide dataset link, I want to practice this code
excellent explanation
This is for mutliclass svm?
Mantap videonya.
Saya juga ada nih rekomendasi lain buat belajar Tuning Hyperparameter KNN pada python 3 siapa tau cocok hehe.
th-cam.com/video/IzwOkGuZpsE/w-d-xo.html