| 1 | A CLASS OF FACTORIZED QUASI-NEWTON METHODS FOR NONLINEAR LEAST SQUARES PROBLEMS显示文摘This paper gives a class of descent methods for nonlinar least squares solu-tion. A class of updating formulae is obtained by using generalized inverse matrices.These formulae generate an approximation to the second part of the Hessian ma-trix of the objective function, and are updated in such a way that the resultingapproximation to the whole Hessian matrix is the convex class of Broyden-like up-dating formulae. It is proved that the proposed updating formulae are invarantunder linear transformation and that the class of factorized quasi-Newton methodsare locally and superlinearly convergent. Numerical results are presented and showthat the proposed methods are promising. | C.X. Xu X.F. Ma M.Y. Kong(Department of Mathematics, Xi’an Jiaotong University, Xi’an, China) | 1996 | Journal of Computational Mathematics1996，14，2： | 4 |
您所在的机构暂未开通此部分的功能，请联系管理员