Tj. Burkholder et Rl. Lieber, STEPWISE REGRESSION IS AN ALTERNATIVE TO SPLINES FOR FITTING NOISY DATA, Journal of biomechanics, 29(2), 1996, pp. 235-238
In this study, we compared numerical methods that are used to fit nois
y data. Comparisons included polynominal regression, stepwise polynomi
al regression and quintic spline approximation. The advantages and lim
itations of each method are discussed in terms of curve fit quality, c
omputational speed and ease, and solution compactness, Overall, the sp
line approximation and stepwise polynomial regression provide the best
fits to the data. Stepwise regression provides the added utility of p
roviding a simple, unconstrained function which can be easily implemen
ted in simulation studies.