The Levenberg-Marquardt method: an overview of modern convergence theories and more

被引:8
|
作者
Fischer, Andreas [1 ]
Izmailov, Alexey F. [2 ,3 ]
Solodov, Mikhail V. [4 ]
机构
[1] Tech Univ Dresden, Fac Math, D-01062 Dresden, Germany
[2] Lomonosov Moscow State Univ MSU, VMK Fac, OR Dept, Uchebniy Korpus 2, Moscow 119991, Russia
[3] Derzhavin Tambov State Univ TSU, Int 33, Tambov 392000, Russia
[4] IMPA Inst Matemat Pura & Aplicada, Estr Dona Castorina 110, BR-22460320 Rio De Janeiro, RJ, Brazil
关键词
Nonlinear equation; Constrained equation; Piecewise smooth equation; Nonisolated solution; Local error bound; Gauss-Newton method; Levenberg-Marquardt method; LP-Newton method; Singular solution; LEAST-SQUARES FORMULATION; NASH EQUILIBRIUM PROBLEMS; NEWTON METHOD; NONLINEAR EQUATIONS; ALGORITHM; SYSTEMS; PARAMETER;
D O I
10.1007/s10589-024-00589-1
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
The Levenberg-Marquardt method is a fundamental regularization technique for the Newton method applied to nonlinear equations, possibly constrained, and possibly with singular or even nonisolated solutions. We review the literature on the subject, in particular relating to each other various convergence frameworks and results. In this process, the analysis is performed from a unified perspective, and some new results are obtained as well. We discuss smooth and piecewise smooth equations, inexact solution of subproblems, and globalization techniques. Attention is also paid to the LP-Newton method, because of its relations to the Levenberg-Marquardt method.
引用
收藏
页码:33 / 67
页数:35
相关论文
共 50 条
  • [21] LEVENBERG-MARQUARDT METHOD WITH A GENERAL LM PARAMETER AND A NONMONOTONE TRUST REGION TECHNIQUE
    Zhao, Luyao
    Tang, Jingyong
    JOURNAL OF APPLIED ANALYSIS AND COMPUTATION, 2024, 14 (04): : 1959 - 1976
  • [22] On the global convergence of a Levenberg-Marquardt method for constrained nonlinear equations
    Yu Z.
    Journal of Applied Mathematics and Computing, 2004, 16 (1-2) : 183 - 194
  • [23] Convergence rate of the modified Levenberg-Marquardt method under Holderian local error bound
    Zheng, Lin
    Chen, Liang
    Tang, Yangxin
    OPEN MATHEMATICS, 2022, 20 (01): : 998 - 1012
  • [24] Convergence of Levenberg-Marquardt method for the inverse problem with an interior measurement
    Jiang, Yu
    Nakamura, Gen
    JOURNAL OF INVERSE AND ILL-POSED PROBLEMS, 2019, 27 (02): : 195 - 215
  • [25] A note on the Levenberg-Marquardt parameter
    Fan, Jinyan
    Pan, Jianyu
    APPLIED MATHEMATICS AND COMPUTATION, 2009, 207 (02) : 351 - 359
  • [26] Some research on Levenberg-Marquardt method for the nonlinear equations
    Ma, Changfeng
    Jiang, Lihua
    APPLIED MATHEMATICS AND COMPUTATION, 2007, 184 (02) : 1032 - 1040
  • [27] A nonsmooth Levenberg-Marquardt method for vertical complementarity problems
    Song, Linsen
    Gao, Yan
    NUMERICAL ALGORITHMS, 2017, 76 (02) : 473 - 485
  • [28] On a New Updating Rule of the Levenberg-Marquardt Parameter
    Zhao, Ruixue
    Fan, Jinyan
    JOURNAL OF SCIENTIFIC COMPUTING, 2018, 74 (02) : 1146 - 1162
  • [29] On the inexactness level of robust Levenberg-Marquardt methods
    Fischer, A.
    Shukla, P. K.
    Wang, M.
    OPTIMIZATION, 2010, 59 (02) : 273 - 287
  • [30] ON QUADRATICAL CONVERGENCE OF INEXACT LEVENBERG-MARQUARDT METHODS UNDER LOCAL ERROR BOUND CONDITION
    Zhang, Yan
    Yu, Carisa Kwok Wai
    Bao, Ji-Feng
    Wang, Jinhua
    JOURNAL OF NONLINEAR AND CONVEX ANALYSIS, 2018, 19 (01) : 123 - 146