Comparison of biasing parameter computational techniques in ridge-type estimation (Q1208325)

Ridge-type estimation methods are often used as a means to overcome the effects of ill-conditioning in linear and nonlinear inverse problems. The most critical aspect in the implementation of these methods is the computation of the biasing parameter that is used to adjust the eigenvalue spectrum of the normal matrix. The computation of the biasing parameter is based upon minimizing the mean squared error of the solution, for which a number of different numerical techniques have been proposed. In this paper a comparison of the performance of proposed biasing parameter computational techniques, both old and new, is presented in the solution of a large-scale space mechanics problem.