Classification guide for Model Parameters
Is there anybody working on a classification guide explaining the meaning and impact of the different Model Parameters of, e.g. SVM classification? Does any material related to this issue exist?
Comments (9)
-
-
reporter and why are you selecting rbf as kernel?
param_grid = {'kernel': ['rbf'],'gamma': [0.001, 0.01, 0.1, 1, 10, 100, 1000],'C': [0.001, 0.01, 0.1, 1, 10, 100, 1000]}
(I know I can change it, just asking why you selected this in the code).
-
Broadly speaking, RBF is more accurate than polynomial. Just do some reading on that, e.g. https://www.researchgate.net/post/Diffference_between_SVM_Linear_polynmial_and_RBF_kernel
-
I guess we could also have a “Fit Polynomial SVC”. Would that be of interest?
-
reporter I think that while rbf is a reasonable default, we should be able of selecting any of the kernels available in scikit-learn in the menu (not having to edit the code, which I find that intimidates many users).
-
Ok, beside RBF and Linear kernel, I will introduce Polynomial as well.
-
- removed component
- changed version to 3.9
-
-
assigned issue to
-
assigned issue to
-
- changed status to resolved
resolves
#621→ <<cset 35fd3cb16c79>>
- Log in to comment
The code snippet used in Classification is pure Scikit-Learn API code: you can read the detailed SVC user guide and class descriptions here:
https://scikit-learn.org/stable/modules/svm.html#classification
https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html#sklearn.svm.SVC