Priority Research Paper Volume 12, Issue 9 pp 7626—7638

Prediction of chronological and biological age from laboratory data

class="figure-viewer-img"

Figure 4. Correlations of feature importance scores across gender and race subgroups. Pairwise correlations between feature importance scores from random forest models trained on subsets of the data (separated by gender and race/ethnicity). Correlations are consistently stronger across race groups for the same gender.