Skip to main content
. 2024 Jun 17;24:75. doi: 10.1186/s40644-024-00719-2

Fig. 4.

Fig. 4

Feature selection using the least absolute shrinkage and selection operator (LASSO) algorithm. (A) A coefficient path plot was generated showing how the coefficients of each variable changed at different regularization levels. (B) AUC (red dots) with standard errors (error bar) can be used to determine the optimum penalty lambda (λ). (C) Boxplot of Radscores between cancer tissues with or without lymph node metastasis from the training set (left) and the test set (right)