Supplementary Figure 7 from A Novel Sensitivity Maximization at a Given Specificity Method for Binary Classifications

作者：Seyyed Mahmood Ghasemi, Chunhui Gu, Johannes F. Fahrmann, Samir Hanash, Kim‐Anh Do, James P. Long, Ehsan Irajizad · 年份：2026 · DOI：10.1158/1940-6207.32703528 · 研究领域：Imbalanced Data Classification Techniques、AI in cancer detection、Statistical Methods in Epidemiology

<p>Supplementary Figure 7 portrays the ROC curves for Logistic Regression and SMAGS, generated for the test set of the colorectal cancer dataset, specifically at a 90% specificity threshold. Each model was trained 100 times using different training sets, each comprising 80% of the entire dataset. After training, the models were tested on the remaining 20%, and True Positive Rate (TPR) and False Positive Rate (FPR) values were recorded. The plot displays the averaged TPR and FPR for both models, with the shaded areas representing the corresponding standard deviations.</p>