Skip to main content
. 2022 Aug 25;13:978865. doi: 10.3389/fimmu.2022.978865

Figure 2.

Figure 2

Predictive performance comparison of the 10 mainstream machine learning algorithms in the TCGA dataset. (A) Distribution of the CRGI and survival status of individual PAAD patients. (B) Survival analysis of the high- and low-CRGI groups. (C) The time-dependent ROC curve with the AUC value of 1-, 2-, 3-, 4-, and 5-year OS prediction of the best model optimized by LASSO penalized Cox regression. (D–L) The predictive accuracy of other machine learning algorithms (i.e., Decision Tree, GMM, GBDT, K-NN, LGBM, Logistic Regression, Random Forest, SVM, and XGboost, respectively).