Tikhonov吉洪诺夫正则化.doc

Tikhonov regularization From Wikipedia, the free encyclopedia Tikhonov regularization is the most commonly used method of  HYPERLINK /wiki/Regularization_(mathematics) \o Regularization (mathematics) regularization of  HYPERLINK /wiki/Ill-posed_problem \o Ill-posed problem ill-posed problems named for  HYPERLINK /wiki/Andrey_Tychonoff \o Andrey Tychonoff Andrey Tychonoff. In  HYPERLINK /wiki/Statistics \o Statistics statistics, the method is also known as ridge regression. It is related to the  HYPERLINK /wiki/Levenberg-Marquardt_algorithm \o Levenberg-Marquardt algorithm Levenberg-Marquardt algorithm for  HYPERLINK /wiki/Non-linear_least_squares \o Non-linear least squares non-linear least-squares problems. The standard approach to solve an  HYPERLINK /wiki/Underdetermined_system \o Underdetermined system underdetermined system of  HYPERLINK /wiki/System_of_linear_equations \o System of linear equations linear equations given as is known as  HYPERLINK /wiki/Linear_least_squares \o Linear least squares linear least squares and seeks to minimize the  HYPERLINK /wiki/Residual_(numerical_analysis) \o Residual (numerical analysis) residual where is the  HYPERLINK /wiki/Norm_(mathematics) \l Euclidean_norm \o Norm (mathematics) Euclidean norm. However, the matrix A may be  HYPERLINK /wiki/Ill-conditioned \o Ill-conditioned ill-conditioned or  HYPERLINK /wiki/Singular_matrix \o Singular matrix singular yielding a non-unique solution. In order to give preference to a particular solution with desirable properties, the regularization term is included in this minimization: for some suitably chosen Tikhonov matrix, . In many cases, this matrix is chosen as the  HYPERLINK /wiki/Identity_matrix \o Identity matrix identity matrix = I, giving preference to solutions with smaller norms. In other cases,  HYPERLINK /wiki/Highpass \o Highpass highpass operators (e.g., a  HYPERLINK /wiki/Difference_operator \o Difference operator differenc