Table 8 Optimal LASSO Regularization Parameters for MMSE and MoCA Prediction Models

From: Classifying the risk of cognitive impairment in Parkinson’s disease using serum bile acid profiles and machine learning

Variables

MMSE

MoCA

lambda.min

0.026

0.028

lambda.1se

0.095

0.087

AUC (95% CI)

0.715 (95% CI 0.608–0.821)

0.737 (95% CI 0.634–0.839)

Sensitivity

0.800

0.533

Specificity

0.556

0.844

  1. lambda.min: the value of the regularization parameter λ that achieves the lowest average cross-validated deviance (or error). lambda.1se: the largest λ whose cross-validated deviance is still within one standard error of the minimum
Back to article page