You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Dec 6, 2023. It is now read-only.
Different estimators have different penalty param names, and also have different meanings:
alpha param in Lasso is a regularization strength (for bigger values of alpha regularization strength is ascending)
C param in LogisticRegression and SVM is a opposition of regularization strength (for bigger values of C regularization strength is descending)
Besides that, plot_stability_path generates different plots for such estimators (shrinking on left or right sides of plot).
Using 1/C syntax it's possible to synchronize these meanings and make it more intuitive.
The text was updated successfully, but these errors were encountered:
a-n-ermakov
changed the title
'1/C' syntax for lambda_name
'1/C' syntax for lambda_name in StabilitySelection class
Mar 15, 2019
a-n-ermakov
changed the title
'1/C' syntax for lambda_name in StabilitySelection class
'1/C' syntax for lambda_name in StabilitySelection class
Mar 16, 2019
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Different estimators have different penalty param names, and also have different meanings:
alpha
param in Lasso is a regularization strength (for bigger values ofalpha
regularization strength is ascending)C
param in LogisticRegression and SVM is a opposition of regularization strength (for bigger values ofC
regularization strength is descending)Besides that,
plot_stability_path
generates different plots for such estimators (shrinking on left or right sides of plot).Using
1/C
syntax it's possible to synchronize these meanings and make it more intuitive.The text was updated successfully, but these errors were encountered: