Machine learning method accuracy from each model with a combination of subareas for the attribution predictor. The order of subareas (1 through 5) reflects the sequential addition and the respective accuracy when included in the model through the use of a greedy algorithm.
datasetposted on 2019-04-15, 17:27 authored by Yu Zhang, Jennifer L. Pechal, Carl J. Schmidt, Heather R. Jordan, Wesley W. Wang, M. Eric Benbow, Sing-Hoi Sze, Aaron M. Tarone
“Subarea” is the microbiota from a specific anatomic area. The model with the highest accuracy within each method for each predictor variable is indicated with an asterisk (*).