Predicting Overall Survival in Patients with Nonmetastatic Gastric Signet Ring Cell Carcinoma: A Machine Learning Approach

Figure 2

A comparison of the area under the receiver operating characteristic curve (AUROC) of different machine learning models in predicting (a) 1-, (b) 3-, and (c) 5-year overall survival (OS) in patients with nonmetastatic GSRC in the training cohort and (d) 1-, (e) 3-, and (f) 5-year OS in the validation cohort. KNN: k-nearest neighbor; SVM: support vector machines; RF: random forest; NT: neural networks; XGB: an eXtreme gradient boosting.

(a)

(b)

(c)

(d)

(e)

(f)