Skip to main content

Table 1 The early lung cancer prediction performances on four different datasets using random forest

From: Genomic analyses based on pulmonary adenocarcinoma in situ reveal early lung cancer signature

Model Assessment TCGA GSE68465 GSE10072 GSE89843 (Blood)
Random Forest Accuracy 98.68% 99.51% 97.91% 64.35%
Sensitivity 99.28% 99.95% 98.05% 78.12%
Specificity 95.68% 92.83% 97.75% 50.59%