Additional file 1 of Helicobacter pylori (H. pylori) risk factor analysis and prevalence prediction: a machine learning-based approach
- Resource Type
- Authors
- Tran, Van; Saad, Tazmilur; Tesfaye, Mehret; Walelign, Sosina; Wordofa, Moges; Abera, Dessie; Desta, Kassu; Tsegaye, Aster; Ay, Ahmet; Taye, Bineyam
- Source
- Subject
- Language
Additional file 1: Figure S1. Average H. pylori prediction accuracy and F1 scores for each classifier and feature selection pair with boosting. Figure S2. Average H. pylori prediction accuracy and F1 score for each classifier and feature selection pair with bagging. Figure S3. The probability of each feature being selected averaged across all ranking-based methods with standard error. Figure S4. H. pylori risk factors' relative importance based on classifier-accuracy and F1-score based SFFS feature selection. Figure S5. The probability of each feature being selected averaged across all subset-based methods with standard error. Figure S6. Average H. pylori prevalence prediction accuracy and F1- scores of machine learning classifiers using various feature selection methods and nested cross validation. Table S1. Confusion matrix for the best performing model in terms of predictive accuracy in the nested cross validation. Figure S7. Average H. pylori prevalence prediction accuracy and F1-scores of machine learning classifiers using various feature selection methods on dataset upscaled with SMOTE.