Machine learning-based modeling to predict “poor/less-than-optimal” vs “good” mobilizer. (A) Schema of machine leaning modeling pipeline to predict “poor/less-than-optimal” vs “good” mobilizer. (B) Tree-based feature selection for linear regression and decision tree modeling, ranked by feature importance. Correlation of age (C), platelet count (D), and BMI (E) with Day 1 PB CD34 count/μL; Day 1 PB CD34 count/μL in normal (BMI: 18.5-25), overweight (BMI: 25-30) and obese (BMI: >30) donors (F); ROC curve of seven machine learning prediction algorithms, including: Decision tree, Linear Regression, Random Forest, Support Vector Machine, Feedforward Neural Networks, AdaBoost and Gradient boosting (G). (H) Model performance evaluated by accuracy, F1 score, and AUC.