Abstract:
We propose and study a new a posteriori method for finding the regularization parameter in M. M. Lavrent'ev's method for solving ill-posed problems. The method is a generalization of the well-known quasioptimum selection of the regularization parameter in A. N. Tykhonov's method. Our approach to finding the regularization parameter in Lavrentiev's method leads to a regular algorithm that is pseudo-optimal with respect to accuracy order.