Skip to main content

Table 2 List of compared methods

From: Performance of binary prediction models in high-correlation low-dimensional settings: a comparison of methods

Method

Abbreviation

Hyperparameters

Logistic regression

LR

–

Lasso penalization

Lasso

λℓ1 (penalty importance)

Ridge penalization

Ridge

λℓ2 (penalty importance)

Elastic Net penalization

ElasticNet

λℓ1, λℓ2 (importance per penalty)

Dropout regularization

Dropout

δ (dropout ratio)

Principal component logistic regression

PCLR

dPCA (number of components)

Linear auto-encoder logistic regression

LAELR

dLAE (number of components)

λLAE (importance of reconstruction loss)

Non-negative logistic regression

LRNN

–