Fig. 1From: Machine learning based refined differential gene expression analysis of pediatric sepsisComparisons of LR, RF100, and XGB100 classifiers evaluated using four different feature selection methods and 10 runs of 10-fold cross-validation experiments. Each boxplot represents the distribution of average AUC score of 10 models evaluated using a given classification algorithm and feature selection method for selecting top k = 1, 2, …, 10 featuresBack to article page