Skip to main content

Table 1 Performance of five multivariate regression methods for the prediction of LEP and TIMP-1: predictions using double cross-validation

From: Evaluation of multiple variate selection methods from a biological perspective: a nutrigenomics case study

Method

LEP

TIMP-1

r 2

SEV

r 2

SEV

LASSO

0.614

1.65

0.698

0.980

ELASTIC NET

0.577

1.75

0.682

0.899

CovProc

0.762

1.31

0.701

0.864

GA

0.451

2.34

0.482

1.242

PLS

0.621

1.63

0.650

0.925

  1. Bold values indicate the best performance
  2. r 2, squared correlation between predicted and actual values; SEV, root mean squared residuals