TABLE 4.

System efficiency comparisons of different machine learning (ML) models.

ML model	R 2	RMSE	MAD
SR	0.95	3.89	2.97
SVR-eps	0.97	2.95	2.22
SVR-nu	0.97	2.94	2.19
RFR	0.99	1.77	1.29
RFR (1–60)	0.99	1.54	1.15

R², coefficient of determination/goodness-of-fit; RMSE, root mean square error; MAD, median absolute deviation; SR, stepwise regression; SVR-eps, support vector regression eps-regression; SVR-nu, support vector regression nu-regression; RFR, random forest regression in the CHS cohort; RFR (1–60), random forest regression at the 1–60 age categories of the CHS cohort.