Skip to main content
. 2017 Aug 28;12(8):e0183518. doi: 10.1371/journal.pone.0183518

Table 6. Estimated coefficients and model error results, for various regularization procedures applied to the prostate data.

The dashed entries correspond to predictors that are estimated “0”.

Variable Lasso Elastic net Alasso CBPR RAL
Intercept 0.635 0.206 2.921 1.844 0.208
lweight −0.282 −0.821 −0.425
age 0.027 0.008 0.029
lbph
svi
lcp 1.135 0.836 1.213 1.266 0.648
gleason 0.088 0.102 0.267
pgg45
lpsa 0.094 0.05 0.374 0.342
Test Error 1.997 1.132 3.112 2.413 0.807