Regularization

The definition of regularization is very broad. In our context, it can be understood as the methods needed to achieve a unique solution to an “ill-posed” inverse problem. This situation typically happens in the case the number of parameters is larger than the number of observations (known information). In general it is always recommended to apply the regularization techniques to solve the inverse problem.

FePEST includes all available regularization techniques of PEST:

• Regularize by Prior Information: Tikhonov

• Subspace Regularization

• Singular Value Decomposition

• Least Squares

• Regularize by Super Parameters

You can find more information of each regularization option in its corresponding section.

Settings

The Regularization section in the Problem Settings dialog provides the complete set of combination of regularization

Regularization options in FePEST