How bad are the BFGS and DFP methods when the objective function is quadratic? View Full Text


Ontology type: schema:ScholarlyArticle     


Article Info

DATE

1986-01

AUTHORS

M. J. D. Powell

ABSTRACT

We study the use of the BFGS and DFP algorithms with step-lengths of one for minimizing quadratic functions of only two variables. The updating formulae in this case imply nonlinear three term recurrence relations between the eigenvalues of consecutive second derivative approximations, which are analysed in order to explain some gross inefficiencies that can occur. Specifically, the BFGS algorithm may require more than 10 iterations to achieve the first decimal place of accuracy, while the performance of the DFP method is far worse. The results help to explain why the DFP method is often less suitable than the BFGS algorithm for general unconstrained optimization calculations, and they show that quadratic functions provide much information about efficiency when the current vector of variables is too far from the solution for an asymptotic convergence analysis. More... »

PAGES

34-47

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/bf01582161

DOI

http://dx.doi.org/10.1007/bf01582161

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1014228284


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0103", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Numerical and Computational Mathematics", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/01", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Mathematical Sciences", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "University of Cambridge", 
          "id": "https://www.grid.ac/institutes/grid.5335.0", 
          "name": [
            "Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Silver Street, CB3 9EW, Cambridge, England"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Powell", 
        "givenName": "M. J. D.", 
        "id": "sg:person.07731545105.07", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07731545105.07"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "https://doi.org/10.1093/comjnl/13.3.317", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1007597686"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "1986-01", 
    "datePublishedReg": "1986-01-01", 
    "description": "We study the use of the BFGS and DFP algorithms with step-lengths of one for minimizing quadratic functions of only two variables. The updating formulae in this case imply nonlinear three term recurrence relations between the eigenvalues of consecutive second derivative approximations, which are analysed in order to explain some gross inefficiencies that can occur. Specifically, the BFGS algorithm may require more than 10 iterations to achieve the first decimal place of accuracy, while the performance of the DFP method is far worse. The results help to explain why the DFP method is often less suitable than the BFGS algorithm for general unconstrained optimization calculations, and they show that quadratic functions provide much information about efficiency when the current vector of variables is too far from the solution for an asymptotic convergence analysis.", 
    "genre": "research_article", 
    "id": "sg:pub.10.1007/bf01582161", 
    "inLanguage": [
      "en"
    ], 
    "isAccessibleForFree": false, 
    "isPartOf": [
      {
        "id": "sg:journal.1047630", 
        "issn": [
          "0025-5610", 
          "1436-4646"
        ], 
        "name": "Mathematical Programming", 
        "type": "Periodical"
      }, 
      {
        "issueNumber": "1", 
        "type": "PublicationIssue"
      }, 
      {
        "type": "PublicationVolume", 
        "volumeNumber": "34"
      }
    ], 
    "name": "How bad are the BFGS and DFP methods when the objective function is quadratic?", 
    "pagination": "34-47", 
    "productId": [
      {
        "name": "readcube_id", 
        "type": "PropertyValue", 
        "value": [
          "43c8c227ef9b4c896a7c02d3f53414bdd4437df05ced9cf190a69699a9047941"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/bf01582161"
        ]
      }, 
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1014228284"
        ]
      }
    ], 
    "sameAs": [
      "https://doi.org/10.1007/bf01582161", 
      "https://app.dimensions.ai/details/publication/pub.1014228284"
    ], 
    "sdDataset": "articles", 
    "sdDatePublished": "2019-04-11T01:11", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-uberresearch-data-dimensions-target-20181106-alternative/cleanup/v134/2549eaecd7973599484d7c17b260dba0a4ecb94b/merge/v9/a6c9fde33151104705d4d7ff012ea9563521a3ce/jats-lookup/v90/0000000001_0000000264/records_8697_00000531.jsonl", 
    "type": "ScholarlyArticle", 
    "url": "http://link.springer.com/10.1007%2FBF01582161"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/bf01582161'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/bf01582161'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/bf01582161'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/bf01582161'


 

This table displays all metadata directly associated to this object as RDF triples.

64 TRIPLES      21 PREDICATES      28 URIs      19 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/bf01582161 schema:about anzsrc-for:01
2 anzsrc-for:0103
3 schema:author N87f47a08dde24e9bba3db1b11b334ec4
4 schema:citation https://doi.org/10.1093/comjnl/13.3.317
5 schema:datePublished 1986-01
6 schema:datePublishedReg 1986-01-01
7 schema:description We study the use of the BFGS and DFP algorithms with step-lengths of one for minimizing quadratic functions of only two variables. The updating formulae in this case imply nonlinear three term recurrence relations between the eigenvalues of consecutive second derivative approximations, which are analysed in order to explain some gross inefficiencies that can occur. Specifically, the BFGS algorithm may require more than 10 iterations to achieve the first decimal place of accuracy, while the performance of the DFP method is far worse. The results help to explain why the DFP method is often less suitable than the BFGS algorithm for general unconstrained optimization calculations, and they show that quadratic functions provide much information about efficiency when the current vector of variables is too far from the solution for an asymptotic convergence analysis.
8 schema:genre research_article
9 schema:inLanguage en
10 schema:isAccessibleForFree false
11 schema:isPartOf N09d0d1cd5cda4c9da7f362f5a40d72fc
12 N26158ddb4e7747a99b3c9680c38e3cec
13 sg:journal.1047630
14 schema:name How bad are the BFGS and DFP methods when the objective function is quadratic?
15 schema:pagination 34-47
16 schema:productId N735966af4ee44e08b9594daac9bcb325
17 Nd5e06696b825481096766950c06a64b1
18 Nda9e1a29499845239129584372f3ec22
19 schema:sameAs https://app.dimensions.ai/details/publication/pub.1014228284
20 https://doi.org/10.1007/bf01582161
21 schema:sdDatePublished 2019-04-11T01:11
22 schema:sdLicense https://scigraph.springernature.com/explorer/license/
23 schema:sdPublisher N1c1fde9ffefa4561819bb75f59521af2
24 schema:url http://link.springer.com/10.1007%2FBF01582161
25 sgo:license sg:explorer/license/
26 sgo:sdDataset articles
27 rdf:type schema:ScholarlyArticle
28 N09d0d1cd5cda4c9da7f362f5a40d72fc schema:volumeNumber 34
29 rdf:type schema:PublicationVolume
30 N1c1fde9ffefa4561819bb75f59521af2 schema:name Springer Nature - SN SciGraph project
31 rdf:type schema:Organization
32 N26158ddb4e7747a99b3c9680c38e3cec schema:issueNumber 1
33 rdf:type schema:PublicationIssue
34 N735966af4ee44e08b9594daac9bcb325 schema:name readcube_id
35 schema:value 43c8c227ef9b4c896a7c02d3f53414bdd4437df05ced9cf190a69699a9047941
36 rdf:type schema:PropertyValue
37 N87f47a08dde24e9bba3db1b11b334ec4 rdf:first sg:person.07731545105.07
38 rdf:rest rdf:nil
39 Nd5e06696b825481096766950c06a64b1 schema:name dimensions_id
40 schema:value pub.1014228284
41 rdf:type schema:PropertyValue
42 Nda9e1a29499845239129584372f3ec22 schema:name doi
43 schema:value 10.1007/bf01582161
44 rdf:type schema:PropertyValue
45 anzsrc-for:01 schema:inDefinedTermSet anzsrc-for:
46 schema:name Mathematical Sciences
47 rdf:type schema:DefinedTerm
48 anzsrc-for:0103 schema:inDefinedTermSet anzsrc-for:
49 schema:name Numerical and Computational Mathematics
50 rdf:type schema:DefinedTerm
51 sg:journal.1047630 schema:issn 0025-5610
52 1436-4646
53 schema:name Mathematical Programming
54 rdf:type schema:Periodical
55 sg:person.07731545105.07 schema:affiliation https://www.grid.ac/institutes/grid.5335.0
56 schema:familyName Powell
57 schema:givenName M. J. D.
58 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07731545105.07
59 rdf:type schema:Person
60 https://doi.org/10.1093/comjnl/13.3.317 schema:sameAs https://app.dimensions.ai/details/publication/pub.1007597686
61 rdf:type schema:CreativeWork
62 https://www.grid.ac/institutes/grid.5335.0 schema:alternateName University of Cambridge
63 schema:name Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Silver Street, CB3 9EW, Cambridge, England
64 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...