Data and Supplementary Materials for “Cost-Sensitive Optimal Classification Tree-Based Method for Lung Cancer Diagnosis”

The study uses de-identified clinical data from health examination participants to develop and evaluate a cost-sensitive optimal classification tree with hyperplanes (COCT-H) for lung cancer diagnosis. The final dataset contains 10,472 records, including 1,309 lung cancer cases and 9,163 non-lung-cancer cases.

The variables cover demographic information, smoking-related information, respiratory disease history, family history of cancer, and laboratory examination indicators. The supplementary materials provide additional information on model hyperparameter settings, data characteristics, and experimental results. These materials are provided to support the verification and reproducibility of the study.