Table 1

Best regression model obtained for the use of cross-sectional data and specified input and output

Cross-sectional data
Given numberInputBest regression (from 19 trained)Output (at age)
1Age, gender, sphereGPR—Sq-exponential
R2: 0.79
RMSE: 0.06 D
AxL/Cr
2Age, gender, AxL/CrGPR—Matern
R2: 0.78
RMSE: 0.87 D
Sphere
  • Axl/Cr, axial length to corneal radius ratio; D, diopter; GPR, Gaussian process regression; R2, coefficient of determination; RMSE, root mean squared error; Sq-exponential, square exponential.