Fig. 3
From: Selection of microbial biomarkers with genetic algorithm and principal component analysis
Frequencies of the species selected from the multiple runs of GA. The GA was run with logistic regression as the classification model and species abundance as the input. GA was used to select the combination of the species for classification of obese individuals from the health group. The number on top of each bar is how many times out of 100 GA runs the corresponding species was selected