A New Incremental Strategy for Function Approximation by Radial Basis Function Neural Networks View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

1999

AUTHORS

A. Esposito , M. Marinaro , D. Oricchio , S. Scarpetta

ABSTRACT

In this paper we present a new incremental learning algorithm for approximating continuous mapping. The algorithm uses radial basis functions as activation functions. The novelty of the algorithm has different aspects: First, the learning procedure is accomplished modifying only the variances of the activation functions instead of the weights on the synapses; Secondly, the variances of the radial basis functions are trained by a two stage strategy, which includes a local optimization of each new neuron variance. An evolutionary optimization algorithm instead of the usual backpropagation algorithm is used to train the variances. The ability of the net to save training time depends on selectively growing the net structure and on the capability of the algorithm to preserve the locality of the activation functions. Moreover, reported are comparisons with other incremental algorithm in the literature. Such comparisons show that our net perform better both in terms of net size (few hidden neuron are sufficient to reach very good approximations), and in terms of computational time (the training phase is accomplished on a PC computer and require few minutes). More... »

PAGES

113-119

References to SciGraph publications

  • 1990-07. Networks and the best approximation property in BIOLOGICAL CYBERNETICS
  • Book

    TITLE

    Neural Nets WIRN VIETRI-98

    ISBN

    978-1-4471-1208-2
    978-1-4471-0811-5

    Author Affiliations

    Identifiers

    URI

    http://scigraph.springernature.com/pub.10.1007/978-1-4471-0811-5_8

    DOI

    http://dx.doi.org/10.1007/978-1-4471-0811-5_8

    DIMENSIONS

    https://app.dimensions.ai/details/publication/pub.1042288095


    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/2008/0801", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Artificial Intelligence and Image Processing", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Information and Computing Sciences", 
            "type": "DefinedTerm"
          }
        ], 
        "author": [
          {
            "affiliation": {
              "name": [
                "International Institute for Advanced Scientific Studies (I.I.A.S.S.), Vietri sul MareSalerno, Italy", 
                "INFM unit\u00e1 di Salerno, Italy"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Esposito", 
            "givenName": "A.", 
            "id": "sg:person.011031612133.55", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.011031612133.55"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Salerno", 
              "id": "https://www.grid.ac/institutes/grid.11780.3f", 
              "name": [
                "International Institute for Advanced Scientific Studies (I.I.A.S.S.), Vietri sul MareSalerno, Italy", 
                "Dipartimento di Fisica Teorica, Universit\u00e1 di Salerno, Via S. Allende, Baronissi, I84081, Salerno, Italy", 
                "INFM unit\u00e1 di Salerno, Italy"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Marinaro", 
            "givenName": "M.", 
            "id": "sg:person.01027564003.17", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01027564003.17"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "name": [
                "International Institute for Advanced Scientific Studies (I.I.A.S.S.), Vietri sul MareSalerno, Italy"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Oricchio", 
            "givenName": "D.", 
            "id": "sg:person.01260542447.00", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01260542447.00"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "University of Salerno", 
              "id": "https://www.grid.ac/institutes/grid.11780.3f", 
              "name": [
                "Dipartimento di Fisica Teorica, Universit\u00e1 di Salerno, Via S. Allende, Baronissi, I84081, Salerno, Italy", 
                "INFM unit\u00e1 di Salerno, Italy"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Scarpetta", 
            "givenName": "S.", 
            "id": "sg:person.01134412304.81", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01134412304.81"
            ], 
            "type": "Person"
          }
        ], 
        "citation": [
          {
            "id": "https://doi.org/10.1162/neco.1989.1.2.281", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1018769097"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/bf00195855", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1020273049", 
              "https://doi.org/10.1007/bf00195855"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/bf00195855", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1020273049", 
              "https://doi.org/10.1007/bf00195855"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "https://doi.org/10.1016/0925-2312(95)00024-0", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1030932400"
            ], 
            "type": "CreativeWork"
          }
        ], 
        "datePublished": "1999", 
        "datePublishedReg": "1999-01-01", 
        "description": "In this paper we present a new incremental learning algorithm for approximating continuous mapping. The algorithm uses radial basis functions as activation functions. The novelty of the algorithm has different aspects: First, the learning procedure is accomplished modifying only the variances of the activation functions instead of the weights on the synapses; Secondly, the variances of the radial basis functions are trained by a two stage strategy, which includes a local optimization of each new neuron variance. An evolutionary optimization algorithm instead of the usual backpropagation algorithm is used to train the variances. The ability of the net to save training time depends on selectively growing the net structure and on the capability of the algorithm to preserve the locality of the activation functions. Moreover, reported are comparisons with other incremental algorithm in the literature. Such comparisons show that our net perform better both in terms of net size (few hidden neuron are sufficient to reach very good approximations), and in terms of computational time (the training phase is accomplished on a PC computer and require few minutes).", 
        "editor": [
          {
            "familyName": "Marinaro", 
            "givenName": "Maria", 
            "type": "Person"
          }, 
          {
            "familyName": "Tagliaferri", 
            "givenName": "Roberto", 
            "type": "Person"
          }
        ], 
        "genre": "chapter", 
        "id": "sg:pub.10.1007/978-1-4471-0811-5_8", 
        "inLanguage": [
          "en"
        ], 
        "isAccessibleForFree": false, 
        "isPartOf": {
          "isbn": [
            "978-1-4471-1208-2", 
            "978-1-4471-0811-5"
          ], 
          "name": "Neural Nets WIRN VIETRI-98", 
          "type": "Book"
        }, 
        "name": "A New Incremental Strategy for Function Approximation by Radial Basis Function Neural Networks", 
        "pagination": "113-119", 
        "productId": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "pub.1042288095"
            ]
          }, 
          {
            "name": "doi", 
            "type": "PropertyValue", 
            "value": [
              "10.1007/978-1-4471-0811-5_8"
            ]
          }, 
          {
            "name": "readcube_id", 
            "type": "PropertyValue", 
            "value": [
              "c970601c65241036e383c0fb80efc8a3f98efdad5f586f72121ce3eef95da065"
            ]
          }
        ], 
        "publisher": {
          "location": "London", 
          "name": "Springer London", 
          "type": "Organisation"
        }, 
        "sameAs": [
          "https://doi.org/10.1007/978-1-4471-0811-5_8", 
          "https://app.dimensions.ai/details/publication/pub.1042288095"
        ], 
        "sdDataset": "chapters", 
        "sdDatePublished": "2019-04-16T09:19", 
        "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/v90/0000000371_0000000371/records_130823_00000003.jsonl", 
        "type": "Chapter", 
        "url": "https://link.springer.com/10.1007%2F978-1-4471-0811-5_8"
      }
    ]
     

    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.10.1007/978-1-4471-0811-5_8'

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

    curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-1-4471-0811-5_8'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-1-4471-0811-5_8'

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

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-1-4471-0811-5_8'


     

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

    108 TRIPLES      23 PREDICATES      30 URIs      20 LITERALS      8 BLANK NODES

    Subject Predicate Object
    1 sg:pub.10.1007/978-1-4471-0811-5_8 schema:about anzsrc-for:08
    2 anzsrc-for:0801
    3 schema:author N83853c85d9d3492bb266d7299b8a46fd
    4 schema:citation sg:pub.10.1007/bf00195855
    5 https://doi.org/10.1016/0925-2312(95)00024-0
    6 https://doi.org/10.1162/neco.1989.1.2.281
    7 schema:datePublished 1999
    8 schema:datePublishedReg 1999-01-01
    9 schema:description In this paper we present a new incremental learning algorithm for approximating continuous mapping. The algorithm uses radial basis functions as activation functions. The novelty of the algorithm has different aspects: First, the learning procedure is accomplished modifying only the variances of the activation functions instead of the weights on the synapses; Secondly, the variances of the radial basis functions are trained by a two stage strategy, which includes a local optimization of each new neuron variance. An evolutionary optimization algorithm instead of the usual backpropagation algorithm is used to train the variances. The ability of the net to save training time depends on selectively growing the net structure and on the capability of the algorithm to preserve the locality of the activation functions. Moreover, reported are comparisons with other incremental algorithm in the literature. Such comparisons show that our net perform better both in terms of net size (few hidden neuron are sufficient to reach very good approximations), and in terms of computational time (the training phase is accomplished on a PC computer and require few minutes).
    10 schema:editor Ncf5ca819010e43539aa3577052eedc31
    11 schema:genre chapter
    12 schema:inLanguage en
    13 schema:isAccessibleForFree false
    14 schema:isPartOf Nb1701a3bf2bb4c78936a53f5d7f64216
    15 schema:name A New Incremental Strategy for Function Approximation by Radial Basis Function Neural Networks
    16 schema:pagination 113-119
    17 schema:productId N18f0e7efe8de41d791b3e2083a6bc61d
    18 Nd0dfae4b90bb4e4cb87acbd3f0b84950
    19 Nfeb22e0bdb5d475899b5dd4c47894cca
    20 schema:publisher Nce580247f1a54ca78cf42309444ab105
    21 schema:sameAs https://app.dimensions.ai/details/publication/pub.1042288095
    22 https://doi.org/10.1007/978-1-4471-0811-5_8
    23 schema:sdDatePublished 2019-04-16T09:19
    24 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    25 schema:sdPublisher Nf3c767d59ba54058bd6d7935476b59e7
    26 schema:url https://link.springer.com/10.1007%2F978-1-4471-0811-5_8
    27 sgo:license sg:explorer/license/
    28 sgo:sdDataset chapters
    29 rdf:type schema:Chapter
    30 N18f0e7efe8de41d791b3e2083a6bc61d schema:name doi
    31 schema:value 10.1007/978-1-4471-0811-5_8
    32 rdf:type schema:PropertyValue
    33 N25295f3c29a245bf9016f0adede30868 rdf:first sg:person.01134412304.81
    34 rdf:rest rdf:nil
    35 N42c370b2580c4bfb87a609350213bc2b schema:name INFM unitá di Salerno, Italy
    36 International Institute for Advanced Scientific Studies (I.I.A.S.S.), Vietri sul MareSalerno, Italy
    37 rdf:type schema:Organization
    38 N451b717c2a93445c9a737c6d2da67cb8 rdf:first sg:person.01260542447.00
    39 rdf:rest N25295f3c29a245bf9016f0adede30868
    40 N49916d9ecf7041e487d8fd42c6cb13b1 rdf:first Nfa5f01a8566f4bf589a83a6ce4ae6ea7
    41 rdf:rest rdf:nil
    42 N83853c85d9d3492bb266d7299b8a46fd rdf:first sg:person.011031612133.55
    43 rdf:rest Nabcb5feb09064fb78126ec3640fa538a
    44 N9739afbdd5dc47a39f05b20cf9840680 schema:name International Institute for Advanced Scientific Studies (I.I.A.S.S.), Vietri sul MareSalerno, Italy
    45 rdf:type schema:Organization
    46 Nabcb5feb09064fb78126ec3640fa538a rdf:first sg:person.01027564003.17
    47 rdf:rest N451b717c2a93445c9a737c6d2da67cb8
    48 Nb1701a3bf2bb4c78936a53f5d7f64216 schema:isbn 978-1-4471-0811-5
    49 978-1-4471-1208-2
    50 schema:name Neural Nets WIRN VIETRI-98
    51 rdf:type schema:Book
    52 Nce580247f1a54ca78cf42309444ab105 schema:location London
    53 schema:name Springer London
    54 rdf:type schema:Organisation
    55 Ncf5ca819010e43539aa3577052eedc31 rdf:first Nf32559aaaaac49f8b75e72624703cd2c
    56 rdf:rest N49916d9ecf7041e487d8fd42c6cb13b1
    57 Nd0dfae4b90bb4e4cb87acbd3f0b84950 schema:name readcube_id
    58 schema:value c970601c65241036e383c0fb80efc8a3f98efdad5f586f72121ce3eef95da065
    59 rdf:type schema:PropertyValue
    60 Nf32559aaaaac49f8b75e72624703cd2c schema:familyName Marinaro
    61 schema:givenName Maria
    62 rdf:type schema:Person
    63 Nf3c767d59ba54058bd6d7935476b59e7 schema:name Springer Nature - SN SciGraph project
    64 rdf:type schema:Organization
    65 Nfa5f01a8566f4bf589a83a6ce4ae6ea7 schema:familyName Tagliaferri
    66 schema:givenName Roberto
    67 rdf:type schema:Person
    68 Nfeb22e0bdb5d475899b5dd4c47894cca schema:name dimensions_id
    69 schema:value pub.1042288095
    70 rdf:type schema:PropertyValue
    71 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
    72 schema:name Information and Computing Sciences
    73 rdf:type schema:DefinedTerm
    74 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
    75 schema:name Artificial Intelligence and Image Processing
    76 rdf:type schema:DefinedTerm
    77 sg:person.01027564003.17 schema:affiliation https://www.grid.ac/institutes/grid.11780.3f
    78 schema:familyName Marinaro
    79 schema:givenName M.
    80 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01027564003.17
    81 rdf:type schema:Person
    82 sg:person.011031612133.55 schema:affiliation N42c370b2580c4bfb87a609350213bc2b
    83 schema:familyName Esposito
    84 schema:givenName A.
    85 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.011031612133.55
    86 rdf:type schema:Person
    87 sg:person.01134412304.81 schema:affiliation https://www.grid.ac/institutes/grid.11780.3f
    88 schema:familyName Scarpetta
    89 schema:givenName S.
    90 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01134412304.81
    91 rdf:type schema:Person
    92 sg:person.01260542447.00 schema:affiliation N9739afbdd5dc47a39f05b20cf9840680
    93 schema:familyName Oricchio
    94 schema:givenName D.
    95 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01260542447.00
    96 rdf:type schema:Person
    97 sg:pub.10.1007/bf00195855 schema:sameAs https://app.dimensions.ai/details/publication/pub.1020273049
    98 https://doi.org/10.1007/bf00195855
    99 rdf:type schema:CreativeWork
    100 https://doi.org/10.1016/0925-2312(95)00024-0 schema:sameAs https://app.dimensions.ai/details/publication/pub.1030932400
    101 rdf:type schema:CreativeWork
    102 https://doi.org/10.1162/neco.1989.1.2.281 schema:sameAs https://app.dimensions.ai/details/publication/pub.1018769097
    103 rdf:type schema:CreativeWork
    104 https://www.grid.ac/institutes/grid.11780.3f schema:alternateName University of Salerno
    105 schema:name Dipartimento di Fisica Teorica, Universitá di Salerno, Via S. Allende, Baronissi, I84081, Salerno, Italy
    106 INFM unitá di Salerno, Italy
    107 International Institute for Advanced Scientific Studies (I.I.A.S.S.), Vietri sul MareSalerno, Italy
    108 rdf:type schema:Organization
     




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


    ...