Skip to main content
[Preprint]. 2024 Sep 9:rs.3.rs-4791069. [Version 1] doi: 10.21203/rs.3.rs-4791069/v1

Figure 5. Machine learning models to classify the HS and Healthy Patients based on the identified biomarkers.

Figure 5.

(A) Top 20 ensemble features predicted using DEGs. (B) Top 20 ensemble features predicted using DEPs. (C) Combined predicted features from DEGs and DEPs. (A-C) Model performance was evaluated using an independent test and accuracy presented in AUC-ROC curves. (D) Accuracies for each cross-validation fold and hyperparameter tuning set are shown for 4 ML models: Gradient Boosting (GB), Neural Network (NN), Random Forest (RF), Support Vector Machine for Classification (SVC).