LASSO回归分析在预后风险预测中的应用

The cleaned data were analyzed using the R package "glmnet" (version 4.1.7) to obtain the lambda value, maximum likelihood number, or C-index, and to visualize the data. Ten-fold cross-validation was employed to screen the LASSO prognostic risk coefficients. The optimal lambda (penalty value) is referred to as lambda.min, and the lambda value within one standard error of the optimal value is referred to as lambda.1se. Additionally, the LASSO variable trajectories were observed to track the changes in the coefficients of variables entering the model.