On the Extension of Constrained Optimization Algorithms from Differentiable to Nondifferentiable Problems
Authors:
- E. PolakEPE. Polak
- D. Q. MayneDMD. Q. Mayne
- Y. WardiYWY. Wardi
- 1 March 1983
- journal article
- Published by Society for Industrial & Applied Mathematics (SIAM) in SIAM Journal on Control and Optimization
- Vol. 21 (2) , 179-203
Abstract
This paper presents three general schemes for extending differentiable optimization algorithms to nondifferentiable problems. It is shown that the Armijo gradient method, phase-I–phase-II methods of feasible directions and exact penalty function methods have conceptual analogs for problems with locally Lipschitz functions and implementable analogs for problems with semismooth functions. The exact penalty method has required the development of a new optimality condition.Keywords
This publication has 34 references indexed in Scilit:
- Optimal Control and the True HamiltonianSIAM Review, 1979
- A New Approach to Lagrange MultipliersMathematics of Operations Research, 1976
- The minimization of certain nondifferentiable sums of eigenvalues of symmetric matrices1975
- Generalized gradients and applicationsTransactions of the American Mathematical Society, 1975
- A Descent Numerical Method for Optimization Problems with Nondifferentiable Cost FunctionalsSIAM Journal on Control, 1973
- Constrained Optimization Using a Nondifferentiable Penalty FunctionSIAM Journal on Numerical Analysis, 1973
- Algorithms for some minimax problemsJournal of Computer and System Sciences, 1968
- Differentiability of a maximum function. IUSSR Computational Mathematics and Mathematical Physics, 1968
- The Theory of Max-Min, with ApplicationsSIAM Journal on Applied Mathematics, 1966
- Minimization of functions having Lipschitz continuous first partial derivativesPacific Journal of Mathematics, 1966