GLMNETHyperparameters

Properties

alphatunable

numberdefault 1≥ 0≤ 1

Elastic net mixing parameter (0 = ridge, 1 = lasso).

family

string | nulldefault null

GLM family. NULL = set from outcome type.

one of"gaussian""binomial""poisson""multinomial""cox""mgaussian"

offset

array | nulldefault null≥ 1 items

Offset. Must have one value per case.

which_lambda_cv

stringdefault "lambda.1se"

Which cross-validated lambda to use for prediction.

one of"lambda.1se""lambda.min"

nlambda

integerdefault 100≥ 1

Number of lambda values.

lambda

array | nulldefault null≥ 1 items

Regularization strength. NULL = determined by cv.glmnet during tuning.

penalty_factor

array | nulldefault null≥ 1 items

Penalty factor, one per feature.

standardize

booleandefault true

Standardize features.

intercept

booleandefault true

Include intercept.

ifwtunable

booleandefault true

Inverse Frequency Weighting in classification.

Relationships