Fig. 3

Feature filtering by LASSO when the dataset was divided into CEN12 versus 34. A Tuning parameter (λ) selection in the LASSO model used tenfold cross-validation via minimum criteria. B LASSO coefficient profiles, A coefficient profile plot was produced against the log (λ) sequence. C Weight map of the final incorporated features, values were the absolute values of the feature coefficients, eight features with the highest absolute values of the coefficients were included for further study