Formulation and Solution of Stochastic Inverse Problems

In Part I, we reviewed the standard econometric enterprise where

1. the origin of observed data is characterized within a sufficiently constrained stochastic model formulation,

2. assumptions are used to define the underlying data sampling process (DSP), and

3. the model contains a sufficiently small number of unknowns so that information about them can be recovered-estimated by long-standing traditional econometric methods.

In this case, the econometric information recovery problem takes on the form of a well-posed just-determined or overdetermined problem where a solution exists, is unique, and there are a sufficient number of data observations to estimate the unknowns.