ABSTRACT
The problem of choosing the smoothing parameter is ubiquitous in curve estimation, even if in certain cases it is swept under the carpet in the way a method is specified. For example, if one is fitting curves by polynomial regression, the choice of the degree of the fitted polynomial is essentially equivalent to the choice of a smoothing parameter. In the spline smoothing methodology set out in Chapter 2 above, the smoothing parameter is of course explicit in the method.
