BFO, a trainable derivative-free Brute Force Optimizer for nonlinear bound-constrained optimization and equilibrium computations with continuous and discrete variables

Margherita Porcelli, Philippe Toint

Research output: Contribution to journalArticle

7 Downloads (Pure)

Abstract

A direct-search derivative-free Matlab optimizer for bound-constrained problems is described, whose remarkable features are its ability to handle a mix of continuous and discrete variables, a versatile interface as well as a novel self-training option. Its performance compares favorably with that of NOMAD (Nonsmooth Optimization by Mesh Adaptive Direct Search), a well-known derivative-free optimization package. It is also applicable to multilevel equilibrium- or constrained-type problems. Its easy-to-use interface provides a number of user-oriented features, such as checkpointing and restart, variable scaling, and early termination tools.

Original languageEnglish
Article number6
Number of pages28
JournalTransactions of the American Methematical Society on Mathematical Software
Volume44
Issue number1
DOIs
Publication statusPublished - 30 Jun 2017

Fingerprint

Direct Search
Derivative-free
Discrete Variables
Constrained optimization
Continuous Variables
Constrained Optimization
Derivative-free Optimization
Early Termination
Derivatives
Checkpointing
Adaptive Mesh
Nonsmooth Optimization
Restart
MATLAB
Scaling
Training

Keywords

  • derivative-free optimization
  • trainable algorithms
  • bound constraints
  • mixed integer optimization
  • direct-search methods

Cite this

@article{b15b6052a0114ac5b61654bcfa85f46e,
title = "BFO, a trainable derivative-free Brute Force Optimizer for nonlinear bound-constrained optimization and equilibrium computations with continuous and discrete variables",
abstract = "A direct-search derivative-free Matlab optimizer for bound-constrained problems is described, whose remarkable features are its ability to handle a mix of continuous and discrete variables, a versatile interface as well as a novel self-training option. Its performance compares favorably with that of NOMAD (Nonsmooth Optimization by Mesh Adaptive Direct Search), a well-known derivative-free optimization package. It is also applicable to multilevel equilibrium- or constrained-type problems. Its easy-to-use interface provides a number of user-oriented features, such as checkpointing and restart, variable scaling, and early termination tools.",
keywords = "derivative-free optimization , trainable algorithms, bound constraints, mixed integer optimization, direct-search methods",
author = "Margherita Porcelli and Philippe Toint",
year = "2017",
month = "6",
day = "30",
doi = "10.1145/3085592",
language = "English",
volume = "44",
journal = "Transactions of the American Methematical Society on Mathematical Software",
issn = "0002-9947",
publisher = "American Mathematical Society",
number = "1",

}

TY - JOUR

T1 - BFO, a trainable derivative-free Brute Force Optimizer for nonlinear bound-constrained optimization and equilibrium computations with continuous and discrete variables

AU - Porcelli, Margherita

AU - Toint, Philippe

PY - 2017/6/30

Y1 - 2017/6/30

N2 - A direct-search derivative-free Matlab optimizer for bound-constrained problems is described, whose remarkable features are its ability to handle a mix of continuous and discrete variables, a versatile interface as well as a novel self-training option. Its performance compares favorably with that of NOMAD (Nonsmooth Optimization by Mesh Adaptive Direct Search), a well-known derivative-free optimization package. It is also applicable to multilevel equilibrium- or constrained-type problems. Its easy-to-use interface provides a number of user-oriented features, such as checkpointing and restart, variable scaling, and early termination tools.

AB - A direct-search derivative-free Matlab optimizer for bound-constrained problems is described, whose remarkable features are its ability to handle a mix of continuous and discrete variables, a versatile interface as well as a novel self-training option. Its performance compares favorably with that of NOMAD (Nonsmooth Optimization by Mesh Adaptive Direct Search), a well-known derivative-free optimization package. It is also applicable to multilevel equilibrium- or constrained-type problems. Its easy-to-use interface provides a number of user-oriented features, such as checkpointing and restart, variable scaling, and early termination tools.

KW - derivative-free optimization

KW - trainable algorithms

KW - bound constraints

KW - mixed integer optimization

KW - direct-search methods

UR - http://www.scopus.com/inward/record.url?scp=85024481256&partnerID=8YFLogxK

U2 - 10.1145/3085592

DO - 10.1145/3085592

M3 - Article

VL - 44

JO - Transactions of the American Methematical Society on Mathematical Software

JF - Transactions of the American Methematical Society on Mathematical Software

SN - 0002-9947

IS - 1

M1 - 6

ER -