Chronological Age Prediction: We introduce a multi-level approach, based solely on body measurements. We developed an eigen body shape model, based on which we perform body shape clustering. Proposed approach resulted in a mean absolute error (MAE) of 5.4 years for NHANES dataset using 10-fold cross validation.
Biological Age Prediction: We introduce a new centroid based approach, based solely on neighborhood information that can be used for human body measurements, biomarkers, and combination of the both.
Surface Based Body Shape Index (SBSI): We introduce a new anthropometric measure SBSI that accounts for both body shape and body size, and evaluate its performance as a predictor of all-cause mortality.