https://github.com/bgranzow/l-bfgs-b

A pure Matlab implementation of L-BFGS-B (LBFGSB)
https://github.com/bgranzow/l-bfgs-b

Last synced: 7 days ago
JSON representation

A pure Matlab implementation of L-BFGS-B (LBFGSB)

• Host: GitHub
• URL: https://github.com/bgranzow/l-bfgs-b
• Owner: bgranzow
• License: mit
• Created: 2016-12-12T22:15:38.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2024-09-11T21:11:11.000Z (4 months ago)
• Last Synced: 2024-11-12T15:43:21.591Z (2 months ago)
• Language: TeX
• Homepage:
• Size: 539 KB
• Stars: 57
• Watchers: 2
• Forks: 15
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
## L-BFGS-B

> A pure Matlab implementation of the L-BFGS-B algorithm.

### Introduction

The L-BFGS-B algorithm is a limited memory quasi-Newton,

gradient based optimzation algorithm to solve problems

of the form:

```

minimize f(x)

such that l <= x <= u

```

### Motivation

The L-BFGS-B algorithm uses a limited memory BFGS

representation of the Hessian matrix, making it well-suited

for optimization problems with a large number of design

variables. Many wrappers (C/C++, Matlab, Python, Julia) to

the [original L-BFGS-B Fortran implementation][1] exist, but a

pure Matlab implementation of the algorithm (as far as I

could tell) did not exist up to this point. This is likely due

to performance concerns. Nevertheless, this single file

implementation (`LBFGS.m`) of the L-BFGS-B algorithm

seeks to provide Matlab users a convenient way to

optimize bound-constrained problems with the L-BFGS-G

algorithm without installing third-party software.

### Background

* The [original L-BFGS-B paper][0]

* The [original L-BFGS-B fortran implementation][1]

* Numerical optimization [fundamentals][2]

[0]:http://epubs.siam.org/doi/abs/10.1137/0916069

[1]:http://users.iems.northwestern.edu/~nocedal/lbfgsb.html

[2]:http://link.springer.com/book/10.1007%2F978-0-387-40065-5

[3]:https://en.wikipedia.org/wiki/Limited-memory_BFGS