Data fitting problems with bounded uncertainties in the data (Q2719198)

The paper is devoted to the data smoothing problem. The main purpose of the study is to show how data fitting with bounded uncertainty problems can be considered in more general setting. In particular, it is shown how such problems can be posed in more computationally convenient form. The special structure is exploited to computational advantage. Some observations are included to develop algorithmic advantages for the particular case, when the underlying norm is the least squares norm. Some variants of the main problems are also considered to show how these problems can be posed in a norm that facilitates their numerical solution.