Derivative free multilevel optimization methods

Download

index.pdf

Date

2009

Author

Pekmen, Bengisen

Metadata

Show full item record

Item Usage Stats

231
views

84
downloads

Derivative free optimization algorithms are implementations of trust region based derivative-free methods using multivariate polynomial interpolation. These are designed to minimize smooth functions whose derivatives are not available or costly to compute. The trust region based multilevel optimization algorithms for solving large scale unconstrained optimization problems resulting by discretization of partial differential equations (PDEs), make use of different discretization levels to reduce the computational cost. In this thesis, a derivative free multilevel optimization algorithm is derived and its convergence behavior is analyzed. The effectiveness of the algorithms is demonstrated on a shape optimization problem.

Subject Keywords

Mathematics.

URI

http://etd.lib.metu.edu.tr/upload/12610853/index.pdf
https://hdl.handle.net/11511/18752

Collections

Graduate School of Applied Mathematics, Thesis

Suggestions

OpenMETU
Core

Noncomplex smooth 4-manifolds with Lefschetz fibrations Korkmaz, Mustafa (2001-01-01) For every integer g ≥ 2 there exist infinitely many pairwise nonhomeomorphic smooth 4-manifolds admitting genus-g Lefschetz fibration over S2 but not carrying any complex structure. This extends a recent result of Ozbagci and Stipsicz.
Non-commutative holomorphic functions in elements of a Lie algebra and the absolute basis problem Dosi (Dosiev), A. A. (IOP Publishing, 2009-11-01) We study the absolute basis problem in algebras of holomorphic functions in non-commuting variables generating a finite-dimensional nilpotent Lie algebra g. This is motivated by J. L. Taylor's programme of non-commutative holomorphic functional calculus in the Lie algebra framework.
Image segmentation and smoothing via partial differential equations Özmen, Neslihan; Leblebicioğlu, Mehmet Kemal; Department of Scientific Computing (2009) In image processing, partial differential equation (PDE) based approaches have been extensively used in segmentation and smoothing applications. The Perona-Malik nonlinear diffusion model is the first PDE based method used in the image smoothing tasks. Afterwards the classical Mumford-Shah model was developed to solve both image segmentation and smoothing problems and it is based on the minimization of an energy functional. It has numerous application areas such as edge detection, motion analysis, medical i...
Dynamic programming for a Markov-switching jump-diffusion Azevedo, N.; Pinheiro, D.; Weber, Gerhard Wilhelm (Elsevier BV, 2014-09-01) We consider an optimal control problem with a deterministic finite horizon and state variable dynamics given by a Markov-switching jump-diffusion stochastic differential equation. Our main results extend the dynamic programming technique to this larger family of stochastic optimal control problems. More specifically, we provide a detailed proof of Bellman's optimality principle (or dynamic programming principle) and obtain the corresponding Hamilton-Jacobi-Belman equation, which turns out to be a partial in...
REFLECTION GROUP RELATIONS ARISING FROM CLUSTER ALGEBRAS Seven, Ahmet İrfan (American Mathematical Society (AMS), 2016-11-01) There is a well-known analogy between cluster algebras and Kac-moody algebras: roughly speaking, Kac-Moody algebras are associated with symmetrizable generalized Cartan matrices while cluster algebras correspond to skew-symmetrizable matrices. In this paper, we study an interplay between these two classes of matrices. We obtain relations in the Weyl groups of Kac-Moody algebras that come from mutation classes of skew-symmetrizable matrices. More precisely, we establish a set of relations satisfied by the re...

Citation Formats

B. Pekmen, “Derivative free multilevel optimization methods,” M.S. - Master of Science, Middle East Technical University, 2009.