2. How do you achieve quality inversion?

2.1 Different components of the estimation error

Whatever the type of inverse problem being addressed, the inverter searches for a parameter vector x (in parameter estimation, otherwise it may be a function x(t)), from data y. The latter may, for example, be measurements (in the case of magnitude estimation and model identification), or constitute a target (in the case of optimal control), or represent the output of a detailed model (in the case of model reduction).

In addition, it has a model represented by an H operator, see § ...