PolyU researchers invent non-invasive diagnostic device Smart-CKD for advancing clinical management of chronic kidney disease (IMAGE)
Caption
The least absolute shrinkage and selection operator (LASSO) algorithm is applied for variable selection from demographic and ultrasound features. (A) The tuning parameter (λ) selection in the LASSO algorithm was determined using 10-fold cross-validation via minimal criteria. The binomial deviation (partial likelihood deviation) curve was plotted against log(λ). The optimal values based on the minimum criteria and one standard error (SE) of the minimum criteria (i.e., the 1-SE criteria) were indicated by the dotted vertical line. For a λ value of 0.102, with log(λ), -2.275 was chosen (1-SE criteria) according to 10-fold cross-validation. (B) Profiles of the LASSO coefficients for the ten variables. A coefficient profile plot was produced against the log(λ) sequence. The dotted vertical line indicated the value selected using 10-fold cross-validation, where the optimal λ resulted in three nonzero coefficients.
Credit
© 2024 Research and Innovation Office, The Hong Kong Polytechnic University. All Rights Reserved.
Usage Restrictions
Nil
License
Original content