Fig. 2

The processes of LASSO regression for screening variables. A The use of tenfold cross-validation to draw vertical lines at selected feature values. B The coefficient profiles of 26 feature variables were obtained from the log (λ) sequence in the LASSO model. Vertical dotted lines are placed at the minimal mean square error (λ = 0.018) and the standard error of the minimum distance (λ = 0.045). Abbreviations: LASSO, least absolute shrinkage and selection operator