Linear ill-posed problems are solvable on the average for all Gaussian measures (Q1345067): Difference between revisions

By an inverse problem the authors understand a pair consisting of a mapping \(A\) from a normed space \(G\) to a normed space \(F_1\) and of a subset \(D\) of \(F_1\). An inverse problem is said to be well-posed if \(AF_1 \subset D\), \(A\) is injective on \(A^{-1} D\), and the inverse of \(A\) on \(D\) is continuous. A problem is called ill-posed if it is not well-posed. The authors introduce a notion of approximation (related to approximation of solutions via computer). A problem which admits arbitrarily close approximation is called solvable. A weaker approximation property leads to weakly-solvable problems. The use of measure theoretical methods makes it possible to introduce the concept of problems well-posed on average. The article provides seven interesting theorems exposing relations among the introduced notions. Explanations linking the mathematics with applications and the computer approach are supplied. Highly readable.