Figure 2.

Selection of clinical features by least absolute shrinkage and selection operator (LASSO) regression and 10-fold cross-validation. (A) A visual plot of the relationship between coefficients for features and the lambda. The coefficient of each feature gradually tended to zero as the lambda value increased and (B) curve of 10-fold cross-validation in the LASSO regression. The dotted vertical line on the left reflects the number of features and optimal log (lambda) corresponding to the smallest mean squared error. With one standard error criterion of optimal log (lambda), the dotted vertical line on the right indicates that the model constructed with 13 variables was relatively accurate and simple (λ = 0.03114428). λ, lambda.