Fig. 2: Performance of different algorithms in the evaluation cohort.
From: Development and validation of a predictive model for extranodal natural killer/T-cell lymphoma
a AUROC for 5-year OS, b c-index for 5-year OS. AUROC area under the receiver operating characteristic curve, OS overall survival, RF random forest, LR logistic regression, SVM support vector machines, kNN k-nearest neighbor, AdaBoost adaptive boosting, Nb naive bayes, XGBoost eXtreme gradient boosting, CoxBoost cox model with likelihood-based boosting, superPC supervised principal component, plsRcox partial least squares regression for cox models, Lasso lasso regression, GBM gradient boosting machine, RSF random survival forest, MLP multilayer perceptron, StepCox stepwise cox regression, TabPFN Tabular Prior-data Fitted Network.
