Serum uric acid and cutaneous involvement in chronic graft-versus-host disease: a real-world exploratory retrospective cohort study

Supplementary Figure 1. Exploratory machine-learning sensitivity analysis. Receiver-operating-characteristic curves from nested cross-validation comparing eight classifiers. These analyses should be treated only as sensitivity analyses and not as evidence of algorithmic superiority. AUC, area under the curve.补充图1. 探索性机器学习敏感性分析。本图呈现了基于嵌套交叉验证得到的受试者工作特征（Receiver-operating characteristic, ROC）曲线，用于对比八种分类器的性能。本类分析仅可作为敏感性分析使用，不应作为判定算法性能优劣的依据。曲线下面积（Area Under the Curve, AUC）。