Sensitivity of Parameter Control Mechanisms with Respect to Their Initialization View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2018-08-21

AUTHORS

Carola Doerr , Markus Wagner

ABSTRACT

The parameter setting problem constitutes one of the major challenges in evolutionary computation, and is subject to considerable research efforts. Since the optimal parameter values can change during the optimization process, efficient parameter control techniques that automatically identify and track reasonable parameter values are sought. A potential drawback of dynamic parameter selection is that state-of-the-art control mechanisms introduces themselves new sets of hyper-parameters, which need to be tuned for the problem at hand. The general hope is that the performance of an algorithm is much less sensitive with respect to these hyper-parameters than with respect to its original parameters. This belief is backed up by a number of empirical and theoretical results. What is less understood in discrete black-box optimization, however, is the influence of the initial parameter value. We contribute with this work an empirical sensitivity analysis for three selected algorithms with self-adjusting parameter choices: the (1 + 1) EA, the 2-rate EA, and the GA. In all three cases we observe fast convergence of the parameters towards their optimal choices. The performance loss of a sub-optimal initialization is shown to be almost negligible for the former two algorithms. For the GA, in contrast, the choice of is more critical; our results suggest to initialize it by a small value. More... »

PAGES

360-372

References to SciGraph publications

Book

TITLE

Parallel Problem Solving from Nature – PPSN XV

ISBN

978-3-319-99258-7
978-3-319-99259-4

From Grant

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-319-99259-4_29

DOI

http://dx.doi.org/10.1007/978-3-319-99259-4_29

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1106254456


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/2008/0102", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Applied 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": "French National Centre for Scientific Research", 
          "id": "https://www.grid.ac/institutes/grid.4444.0", 
          "name": [
            "Sorbonne Universit\u00e9, CNRS, Laboratoire d\u2019Informatique de Paris 6, LIP6, 75005, Paris, France"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Doerr", 
        "givenName": "Carola", 
        "id": "sg:person.010360414373.45", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010360414373.45"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of Adelaide", 
          "id": "https://www.grid.ac/institutes/grid.1010.0", 
          "name": [
            "Optimisation and Logistics, University of Adelaide, 5005, Adelaide, SA, Australia"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Wagner", 
        "givenName": "Markus", 
        "id": "sg:person.012277443734.79", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012277443734.79"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "sg:pub.10.1007/978-3-642-15844-5_1", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1000452291", 
          "https://doi.org/10.1007/978-3-642-15844-5_1"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1016/j.orp.2016.09.002", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1005297347"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1016/j.tcs.2014.11.028", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1008296423"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1145/2908812.2908950", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1017940419"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1145/1570256.1570306", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1031472048"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-3-540-69432-8", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1039448212", 
          "https://doi.org/10.1007/978-3-540-69432-8"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-3-540-69432-8", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1039448212", 
          "https://doi.org/10.1007/978-3-540-69432-8"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1162/106365601750190398", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1043473749"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1145/2996355", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1047025530"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/978-3-642-25566-3_40", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1052722418", 
          "https://doi.org/10.1007/978-3-642-25566-3_40"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/4235.771166", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061172023"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/tac.1968.1098903", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061469715"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/tevc.2014.2308294", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061605203"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s00453-017-0354-9", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1091088346", 
          "https://doi.org/10.1007/s00453-017-0354-9"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/s00453-017-0354-9", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1091088346", 
          "https://doi.org/10.1007/s00453-017-0354-9"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1613/jair.2861", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1105674394"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "2018-08-21", 
    "datePublishedReg": "2018-08-21", 
    "description": "The parameter setting problem constitutes one of the major challenges in evolutionary computation, and is subject to considerable research efforts. Since the optimal parameter values can change during the optimization process, efficient parameter control techniques that automatically identify and track reasonable parameter values are sought. A potential drawback of dynamic parameter selection is that state-of-the-art control mechanisms introduces themselves new sets of hyper-parameters, which need to be tuned for the problem at hand. The general hope is that the performance of an algorithm is much less sensitive with respect to these hyper-parameters than with respect to its original parameters. This belief is backed up by a number of empirical and theoretical results. What is less understood in discrete black-box optimization, however, is the influence of the initial parameter value. We contribute with this work an empirical sensitivity analysis for three selected algorithms with self-adjusting parameter choices: the (1 + 1) EA, the 2-rate EA, and the GA. In all three cases we observe fast convergence of the parameters towards their optimal choices. The performance loss of a sub-optimal initialization is shown to be almost negligible for the former two algorithms. For the GA, in contrast, the choice of is more critical; our results suggest to initialize it by a small value.", 
    "editor": [
      {
        "familyName": "Auger", 
        "givenName": "Anne", 
        "type": "Person"
      }, 
      {
        "familyName": "Fonseca", 
        "givenName": "Carlos M.", 
        "type": "Person"
      }, 
      {
        "familyName": "Louren\u00e7o", 
        "givenName": "Nuno", 
        "type": "Person"
      }, 
      {
        "familyName": "Machado", 
        "givenName": "Penousal", 
        "type": "Person"
      }, 
      {
        "familyName": "Paquete", 
        "givenName": "Lu\u00eds", 
        "type": "Person"
      }, 
      {
        "familyName": "Whitley", 
        "givenName": "Darrell", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-319-99259-4_29", 
    "inLanguage": [
      "en"
    ], 
    "isAccessibleForFree": false, 
    "isFundedItemOf": [
      {
        "id": "sg:grant.5127107", 
        "type": "MonetaryGrant"
      }
    ], 
    "isPartOf": {
      "isbn": [
        "978-3-319-99258-7", 
        "978-3-319-99259-4"
      ], 
      "name": "Parallel Problem Solving from Nature \u2013 PPSN XV", 
      "type": "Book"
    }, 
    "name": "Sensitivity of Parameter Control Mechanisms with Respect to Their Initialization", 
    "pagination": "360-372", 
    "productId": [
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-319-99259-4_29"
        ]
      }, 
      {
        "name": "readcube_id", 
        "type": "PropertyValue", 
        "value": [
          "12ad9928ddd04d65508a85059e6be4e22a04f6f17e999b5f0af1d8202d0ad59c"
        ]
      }, 
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1106254456"
        ]
      }
    ], 
    "publisher": {
      "location": "Cham", 
      "name": "Springer International Publishing", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-319-99259-4_29", 
      "https://app.dimensions.ai/details/publication/pub.1106254456"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2019-04-16T05:00", 
    "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/v90/0000000325_0000000325/records_100797_00000000.jsonl", 
    "type": "Chapter", 
    "url": "https://link.springer.com/10.1007%2F978-3-319-99259-4_29"
  }
]
 

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.10.1007/978-3-319-99259-4_29'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-319-99259-4_29'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-319-99259-4_29'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-319-99259-4_29'


 

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

148 TRIPLES      23 PREDICATES      40 URIs      19 LITERALS      8 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-319-99259-4_29 schema:about anzsrc-for:01
2 anzsrc-for:0102
3 schema:author N03a9d4110d0f459d94d55894a112baa9
4 schema:citation sg:pub.10.1007/978-3-540-69432-8
5 sg:pub.10.1007/978-3-642-15844-5_1
6 sg:pub.10.1007/978-3-642-25566-3_40
7 sg:pub.10.1007/s00453-017-0354-9
8 https://doi.org/10.1016/j.orp.2016.09.002
9 https://doi.org/10.1016/j.tcs.2014.11.028
10 https://doi.org/10.1109/4235.771166
11 https://doi.org/10.1109/tac.1968.1098903
12 https://doi.org/10.1109/tevc.2014.2308294
13 https://doi.org/10.1145/1570256.1570306
14 https://doi.org/10.1145/2908812.2908950
15 https://doi.org/10.1145/2996355
16 https://doi.org/10.1162/106365601750190398
17 https://doi.org/10.1613/jair.2861
18 schema:datePublished 2018-08-21
19 schema:datePublishedReg 2018-08-21
20 schema:description The parameter setting problem constitutes one of the major challenges in evolutionary computation, and is subject to considerable research efforts. Since the optimal parameter values can change during the optimization process, efficient parameter control techniques that automatically identify and track reasonable parameter values are sought. A potential drawback of dynamic parameter selection is that state-of-the-art control mechanisms introduces themselves new sets of hyper-parameters, which need to be tuned for the problem at hand. The general hope is that the performance of an algorithm is much less sensitive with respect to these hyper-parameters than with respect to its original parameters. This belief is backed up by a number of empirical and theoretical results. What is less understood in discrete black-box optimization, however, is the influence of the initial parameter value. We contribute with this work an empirical sensitivity analysis for three selected algorithms with self-adjusting parameter choices: the (1 + 1) EA, the 2-rate EA, and the GA. In all three cases we observe fast convergence of the parameters towards their optimal choices. The performance loss of a sub-optimal initialization is shown to be almost negligible for the former two algorithms. For the GA, in contrast, the choice of is more critical; our results suggest to initialize it by a small value.
21 schema:editor N70dc3532769d42d9a9b96de9245c813b
22 schema:genre chapter
23 schema:inLanguage en
24 schema:isAccessibleForFree false
25 schema:isPartOf N0fc3f3e262864582893b9d1aff083ee2
26 schema:name Sensitivity of Parameter Control Mechanisms with Respect to Their Initialization
27 schema:pagination 360-372
28 schema:productId N5446476936a74193a8b3e758b70f2118
29 N63add624793e4443a499000d34eeea66
30 N71360a04edf4433ca70e449b204bb446
31 schema:publisher N6019a8e2f990461bad0acfae5c5b7e51
32 schema:sameAs https://app.dimensions.ai/details/publication/pub.1106254456
33 https://doi.org/10.1007/978-3-319-99259-4_29
34 schema:sdDatePublished 2019-04-16T05:00
35 schema:sdLicense https://scigraph.springernature.com/explorer/license/
36 schema:sdPublisher Nb9c44fbafd8b43c987d3f37da9155066
37 schema:url https://link.springer.com/10.1007%2F978-3-319-99259-4_29
38 sgo:license sg:explorer/license/
39 sgo:sdDataset chapters
40 rdf:type schema:Chapter
41 N01b3acc005a347a3bb1cde0c2628a343 schema:familyName Whitley
42 schema:givenName Darrell
43 rdf:type schema:Person
44 N03a9d4110d0f459d94d55894a112baa9 rdf:first sg:person.010360414373.45
45 rdf:rest N353c05275dd449fe90f4150a836e3efd
46 N0fc3f3e262864582893b9d1aff083ee2 schema:isbn 978-3-319-99258-7
47 978-3-319-99259-4
48 schema:name Parallel Problem Solving from Nature – PPSN XV
49 rdf:type schema:Book
50 N177ad283ae6d444696914a42cadf7a72 rdf:first Na9c52411e8bd45a69ade6c344de730b4
51 rdf:rest N457b642847c346e89f428b7334c794d9
52 N1d93416310d94dce97851eb80ca3b9ac schema:familyName Lourenço
53 schema:givenName Nuno
54 rdf:type schema:Person
55 N353c05275dd449fe90f4150a836e3efd rdf:first sg:person.012277443734.79
56 rdf:rest rdf:nil
57 N36af3e6ebaa54f2ab7ba9c5b1864420c rdf:first Nc80a1751b64249af81949c859acdf2f7
58 rdf:rest N7695718967f84125b1e5c93f3370ad9c
59 N457b642847c346e89f428b7334c794d9 rdf:first Nf4d10cb68cd0421aa72365669cd92eb8
60 rdf:rest N93543d392e734c05be32db23244e783f
61 N5446476936a74193a8b3e758b70f2118 schema:name dimensions_id
62 schema:value pub.1106254456
63 rdf:type schema:PropertyValue
64 N6019a8e2f990461bad0acfae5c5b7e51 schema:location Cham
65 schema:name Springer International Publishing
66 rdf:type schema:Organisation
67 N63add624793e4443a499000d34eeea66 schema:name readcube_id
68 schema:value 12ad9928ddd04d65508a85059e6be4e22a04f6f17e999b5f0af1d8202d0ad59c
69 rdf:type schema:PropertyValue
70 N67a611c6cf8249719f8b1e40ec9b6e2f schema:familyName Auger
71 schema:givenName Anne
72 rdf:type schema:Person
73 N70dc3532769d42d9a9b96de9245c813b rdf:first N67a611c6cf8249719f8b1e40ec9b6e2f
74 rdf:rest N36af3e6ebaa54f2ab7ba9c5b1864420c
75 N71360a04edf4433ca70e449b204bb446 schema:name doi
76 schema:value 10.1007/978-3-319-99259-4_29
77 rdf:type schema:PropertyValue
78 N7695718967f84125b1e5c93f3370ad9c rdf:first N1d93416310d94dce97851eb80ca3b9ac
79 rdf:rest N177ad283ae6d444696914a42cadf7a72
80 N93543d392e734c05be32db23244e783f rdf:first N01b3acc005a347a3bb1cde0c2628a343
81 rdf:rest rdf:nil
82 Na9c52411e8bd45a69ade6c344de730b4 schema:familyName Machado
83 schema:givenName Penousal
84 rdf:type schema:Person
85 Nb9c44fbafd8b43c987d3f37da9155066 schema:name Springer Nature - SN SciGraph project
86 rdf:type schema:Organization
87 Nc80a1751b64249af81949c859acdf2f7 schema:familyName Fonseca
88 schema:givenName Carlos M.
89 rdf:type schema:Person
90 Nf4d10cb68cd0421aa72365669cd92eb8 schema:familyName Paquete
91 schema:givenName Luís
92 rdf:type schema:Person
93 anzsrc-for:01 schema:inDefinedTermSet anzsrc-for:
94 schema:name Mathematical Sciences
95 rdf:type schema:DefinedTerm
96 anzsrc-for:0102 schema:inDefinedTermSet anzsrc-for:
97 schema:name Applied Mathematics
98 rdf:type schema:DefinedTerm
99 sg:grant.5127107 http://pending.schema.org/fundedItem sg:pub.10.1007/978-3-319-99259-4_29
100 rdf:type schema:MonetaryGrant
101 sg:person.010360414373.45 schema:affiliation https://www.grid.ac/institutes/grid.4444.0
102 schema:familyName Doerr
103 schema:givenName Carola
104 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.010360414373.45
105 rdf:type schema:Person
106 sg:person.012277443734.79 schema:affiliation https://www.grid.ac/institutes/grid.1010.0
107 schema:familyName Wagner
108 schema:givenName Markus
109 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012277443734.79
110 rdf:type schema:Person
111 sg:pub.10.1007/978-3-540-69432-8 schema:sameAs https://app.dimensions.ai/details/publication/pub.1039448212
112 https://doi.org/10.1007/978-3-540-69432-8
113 rdf:type schema:CreativeWork
114 sg:pub.10.1007/978-3-642-15844-5_1 schema:sameAs https://app.dimensions.ai/details/publication/pub.1000452291
115 https://doi.org/10.1007/978-3-642-15844-5_1
116 rdf:type schema:CreativeWork
117 sg:pub.10.1007/978-3-642-25566-3_40 schema:sameAs https://app.dimensions.ai/details/publication/pub.1052722418
118 https://doi.org/10.1007/978-3-642-25566-3_40
119 rdf:type schema:CreativeWork
120 sg:pub.10.1007/s00453-017-0354-9 schema:sameAs https://app.dimensions.ai/details/publication/pub.1091088346
121 https://doi.org/10.1007/s00453-017-0354-9
122 rdf:type schema:CreativeWork
123 https://doi.org/10.1016/j.orp.2016.09.002 schema:sameAs https://app.dimensions.ai/details/publication/pub.1005297347
124 rdf:type schema:CreativeWork
125 https://doi.org/10.1016/j.tcs.2014.11.028 schema:sameAs https://app.dimensions.ai/details/publication/pub.1008296423
126 rdf:type schema:CreativeWork
127 https://doi.org/10.1109/4235.771166 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061172023
128 rdf:type schema:CreativeWork
129 https://doi.org/10.1109/tac.1968.1098903 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061469715
130 rdf:type schema:CreativeWork
131 https://doi.org/10.1109/tevc.2014.2308294 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061605203
132 rdf:type schema:CreativeWork
133 https://doi.org/10.1145/1570256.1570306 schema:sameAs https://app.dimensions.ai/details/publication/pub.1031472048
134 rdf:type schema:CreativeWork
135 https://doi.org/10.1145/2908812.2908950 schema:sameAs https://app.dimensions.ai/details/publication/pub.1017940419
136 rdf:type schema:CreativeWork
137 https://doi.org/10.1145/2996355 schema:sameAs https://app.dimensions.ai/details/publication/pub.1047025530
138 rdf:type schema:CreativeWork
139 https://doi.org/10.1162/106365601750190398 schema:sameAs https://app.dimensions.ai/details/publication/pub.1043473749
140 rdf:type schema:CreativeWork
141 https://doi.org/10.1613/jair.2861 schema:sameAs https://app.dimensions.ai/details/publication/pub.1105674394
142 rdf:type schema:CreativeWork
143 https://www.grid.ac/institutes/grid.1010.0 schema:alternateName University of Adelaide
144 schema:name Optimisation and Logistics, University of Adelaide, 5005, Adelaide, SA, Australia
145 rdf:type schema:Organization
146 https://www.grid.ac/institutes/grid.4444.0 schema:alternateName French National Centre for Scientific Research
147 schema:name Sorbonne Université, CNRS, Laboratoire d’Informatique de Paris 6, LIP6, 75005, Paris, France
148 rdf:type schema:Organization
 




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


...