Penalized Least Squares Tikhonov Regularization Spline Approximation
Institute of Mathematics and Informatics Bulgarian Academy of Sciences
Serdica Mathematical Journal, Vol. 28, No 4, (2002), 329p-348p
Bounds on the error of certain penalized least squares data
fitting methods are derived. In addition to general results in a fairly abstract
setting, more detailed results are included for several particularly interesting
special cases, including splines in both one and several variables.
* Supported by the Army Research Office under grant DAAD-19-02-10059.