A comparative evaluation of stochastic-based inference methods for Gaussian process models View Full Text


Ontology type: schema:ScholarlyArticle      Open Access: True


Article Info

DATE

2013-06-28

AUTHORS

M. Filippone, M. Zhong, M. Girolami

ABSTRACT

Gaussian Process (GP) models are extensively used in data analysis given their flexible modeling capabilities and interpretability. The fully Bayesian treatment of GP models is analytically intractable, and therefore it is necessary to resort to either deterministic or stochastic approximations. This paper focuses on stochastic-based inference techniques. After discussing the challenges associated with the fully Bayesian treatment of GP models, a number of inference strategies based on Markov chain Monte Carlo methods are presented and rigorously assessed. In particular, strategies based on efficient parameterizations and efficient proposal mechanisms are extensively compared on simulated and real data on the basis of convergence speed, sampling efficiency, and computational cost. More... »

PAGES

93-114

References to SciGraph publications

  • 1996. Bayesian Learning for Neural Networks in NONE
  • 2002-12. Langevin Diffusions and Metropolis-Hastings Algorithms in METHODOLOGY AND COMPUTING IN APPLIED PROBABILITY
  • Identifiers

    URI

    http://scigraph.springernature.com/pub.10.1007/s10994-013-5388-x

    DOI

    http://dx.doi.org/10.1007/s10994-013-5388-x

    DIMENSIONS

    https://app.dimensions.ai/details/publication/pub.1032695104


    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/2008/08", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Information and Computing Sciences", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/17", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Psychology and Cognitive Sciences", 
            "type": "DefinedTerm"
          }, 
          {
            "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/2008/0806", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Information Systems", 
            "type": "DefinedTerm"
          }, 
          {
            "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/1702", 
            "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
            "name": "Cognitive Sciences", 
            "type": "DefinedTerm"
          }
        ], 
        "author": [
          {
            "affiliation": {
              "alternateName": "School of Computing Science, University of Glasgow, Glasgow, UK", 
              "id": "http://www.grid.ac/institutes/grid.8756.c", 
              "name": [
                "School of Computing Science, University of Glasgow, Glasgow, UK"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Filippone", 
            "givenName": "M.", 
            "id": "sg:person.07706215665.03", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07706215665.03"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Department of Biomedical Engineering, Dalian University of Technology, Dalian, P.R. China", 
              "id": "http://www.grid.ac/institutes/grid.30055.33", 
              "name": [
                "Department of Biomedical Engineering, Dalian University of Technology, Dalian, P.R. China"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Zhong", 
            "givenName": "M.", 
            "id": "sg:person.012215654253.40", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012215654253.40"
            ], 
            "type": "Person"
          }, 
          {
            "affiliation": {
              "alternateName": "Department of Statistical Science, University College London, London, UK", 
              "id": "http://www.grid.ac/institutes/grid.83440.3b", 
              "name": [
                "Department of Statistical Science, University College London, London, UK"
              ], 
              "type": "Organization"
            }, 
            "familyName": "Girolami", 
            "givenName": "M.", 
            "id": "sg:person.01206735310.03", 
            "sameAs": [
              "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01206735310.03"
            ], 
            "type": "Person"
          }
        ], 
        "citation": [
          {
            "id": "sg:pub.10.1023/a:1023562417138", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1036982513", 
              "https://doi.org/10.1023/a:1023562417138"
            ], 
            "type": "CreativeWork"
          }, 
          {
            "id": "sg:pub.10.1007/978-1-4612-0745-0", 
            "sameAs": [
              "https://app.dimensions.ai/details/publication/pub.1052635697", 
              "https://doi.org/10.1007/978-1-4612-0745-0"
            ], 
            "type": "CreativeWork"
          }
        ], 
        "datePublished": "2013-06-28", 
        "datePublishedReg": "2013-06-28", 
        "description": "Gaussian Process (GP) models are extensively used in data analysis given their flexible modeling capabilities and interpretability. The fully Bayesian treatment of GP models is analytically intractable, and therefore it is necessary to resort to either deterministic or stochastic approximations. This paper focuses on stochastic-based inference techniques. After discussing the challenges associated with the fully Bayesian treatment of GP models, a number of inference strategies based on Markov chain Monte Carlo methods are presented and rigorously assessed. In particular, strategies based on efficient parameterizations and efficient proposal mechanisms are extensively compared on simulated and real data on the basis of convergence speed, sampling efficiency, and computational cost.", 
        "genre": "article", 
        "id": "sg:pub.10.1007/s10994-013-5388-x", 
        "inLanguage": "en", 
        "isAccessibleForFree": true, 
        "isFundedItemOf": [
          {
            "id": "sg:grant.2787267", 
            "type": "MonetaryGrant"
          }
        ], 
        "isPartOf": [
          {
            "id": "sg:journal.1125588", 
            "issn": [
              "0885-6125", 
              "1573-0565"
            ], 
            "name": "Machine Learning", 
            "publisher": "Springer Nature", 
            "type": "Periodical"
          }, 
          {
            "issueNumber": "1", 
            "type": "PublicationIssue"
          }, 
          {
            "type": "PublicationVolume", 
            "volumeNumber": "93"
          }
        ], 
        "keywords": [
          "Gaussian process model", 
          "Bayesian treatment", 
          "Markov chain Monte Carlo methods", 
          "GP model", 
          "Monte Carlo method", 
          "flexible modeling capability", 
          "stochastic approximation", 
          "process model", 
          "Carlo method", 
          "inference methods", 
          "proposal mechanism", 
          "convergence speed", 
          "efficient parameterization", 
          "inference techniques", 
          "computational cost", 
          "real data", 
          "modeling capabilities", 
          "inference strategy", 
          "sampling efficiency", 
          "data analysis", 
          "model", 
          "approximation", 
          "interpretability", 
          "parameterization", 
          "speed", 
          "technique", 
          "cost", 
          "efficiency", 
          "strategies", 
          "number", 
          "capability", 
          "comparative evaluation", 
          "analysis", 
          "data", 
          "basis", 
          "challenges", 
          "evaluation", 
          "mechanism", 
          "treatment", 
          "method", 
          "paper", 
          "stochastic-based inference techniques", 
          "chain Monte Carlo methods", 
          "efficient proposal mechanisms", 
          "stochastic-based inference methods"
        ], 
        "name": "A comparative evaluation of stochastic-based inference methods for Gaussian process models", 
        "pagination": "93-114", 
        "productId": [
          {
            "name": "dimensions_id", 
            "type": "PropertyValue", 
            "value": [
              "pub.1032695104"
            ]
          }, 
          {
            "name": "doi", 
            "type": "PropertyValue", 
            "value": [
              "10.1007/s10994-013-5388-x"
            ]
          }
        ], 
        "sameAs": [
          "https://doi.org/10.1007/s10994-013-5388-x", 
          "https://app.dimensions.ai/details/publication/pub.1032695104"
        ], 
        "sdDataset": "articles", 
        "sdDatePublished": "2022-01-01T18:30", 
        "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
        "sdPublisher": {
          "name": "Springer Nature - SN SciGraph project", 
          "type": "Organization"
        }, 
        "sdSource": "s3://com-springernature-scigraph/baseset/20220101/entities/gbq_results/article/article_598.jsonl", 
        "type": "ScholarlyArticle", 
        "url": "https://doi.org/10.1007/s10994-013-5388-x"
      }
    ]
     

    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/s10994-013-5388-x'

    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/s10994-013-5388-x'

    Turtle is a human-readable linked data format.

    curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/s10994-013-5388-x'

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

    curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/s10994-013-5388-x'


     

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

    145 TRIPLES      22 PREDICATES      75 URIs      62 LITERALS      6 BLANK NODES

    Subject Predicate Object
    1 sg:pub.10.1007/s10994-013-5388-x schema:about anzsrc-for:08
    2 anzsrc-for:0801
    3 anzsrc-for:0806
    4 anzsrc-for:17
    5 anzsrc-for:1702
    6 schema:author N795ef1dcd4df441a887eb04c84bf0b80
    7 schema:citation sg:pub.10.1007/978-1-4612-0745-0
    8 sg:pub.10.1023/a:1023562417138
    9 schema:datePublished 2013-06-28
    10 schema:datePublishedReg 2013-06-28
    11 schema:description Gaussian Process (GP) models are extensively used in data analysis given their flexible modeling capabilities and interpretability. The fully Bayesian treatment of GP models is analytically intractable, and therefore it is necessary to resort to either deterministic or stochastic approximations. This paper focuses on stochastic-based inference techniques. After discussing the challenges associated with the fully Bayesian treatment of GP models, a number of inference strategies based on Markov chain Monte Carlo methods are presented and rigorously assessed. In particular, strategies based on efficient parameterizations and efficient proposal mechanisms are extensively compared on simulated and real data on the basis of convergence speed, sampling efficiency, and computational cost.
    12 schema:genre article
    13 schema:inLanguage en
    14 schema:isAccessibleForFree true
    15 schema:isPartOf N197a218753a7446ea0301fafc7422e5e
    16 N2d1e01bd602746c78778c9dd5078973e
    17 sg:journal.1125588
    18 schema:keywords Bayesian treatment
    19 Carlo method
    20 GP model
    21 Gaussian process model
    22 Markov chain Monte Carlo methods
    23 Monte Carlo method
    24 analysis
    25 approximation
    26 basis
    27 capability
    28 chain Monte Carlo methods
    29 challenges
    30 comparative evaluation
    31 computational cost
    32 convergence speed
    33 cost
    34 data
    35 data analysis
    36 efficiency
    37 efficient parameterization
    38 efficient proposal mechanisms
    39 evaluation
    40 flexible modeling capability
    41 inference methods
    42 inference strategy
    43 inference techniques
    44 interpretability
    45 mechanism
    46 method
    47 model
    48 modeling capabilities
    49 number
    50 paper
    51 parameterization
    52 process model
    53 proposal mechanism
    54 real data
    55 sampling efficiency
    56 speed
    57 stochastic approximation
    58 stochastic-based inference methods
    59 stochastic-based inference techniques
    60 strategies
    61 technique
    62 treatment
    63 schema:name A comparative evaluation of stochastic-based inference methods for Gaussian process models
    64 schema:pagination 93-114
    65 schema:productId N35e11f13fe2f409795652419e28aa2ae
    66 N687016e9ac97461b94b6f32b06bf3e32
    67 schema:sameAs https://app.dimensions.ai/details/publication/pub.1032695104
    68 https://doi.org/10.1007/s10994-013-5388-x
    69 schema:sdDatePublished 2022-01-01T18:30
    70 schema:sdLicense https://scigraph.springernature.com/explorer/license/
    71 schema:sdPublisher Nfc9424fc5ee44d6eb4dec49bdfc9e81d
    72 schema:url https://doi.org/10.1007/s10994-013-5388-x
    73 sgo:license sg:explorer/license/
    74 sgo:sdDataset articles
    75 rdf:type schema:ScholarlyArticle
    76 N197a218753a7446ea0301fafc7422e5e schema:volumeNumber 93
    77 rdf:type schema:PublicationVolume
    78 N1fe81891198842e7baa13022fd358136 rdf:first sg:person.01206735310.03
    79 rdf:rest rdf:nil
    80 N2d1e01bd602746c78778c9dd5078973e schema:issueNumber 1
    81 rdf:type schema:PublicationIssue
    82 N35e11f13fe2f409795652419e28aa2ae schema:name doi
    83 schema:value 10.1007/s10994-013-5388-x
    84 rdf:type schema:PropertyValue
    85 N687016e9ac97461b94b6f32b06bf3e32 schema:name dimensions_id
    86 schema:value pub.1032695104
    87 rdf:type schema:PropertyValue
    88 N795ef1dcd4df441a887eb04c84bf0b80 rdf:first sg:person.07706215665.03
    89 rdf:rest N8edfc1286d654f14806623265b2b03e3
    90 N8edfc1286d654f14806623265b2b03e3 rdf:first sg:person.012215654253.40
    91 rdf:rest N1fe81891198842e7baa13022fd358136
    92 Nfc9424fc5ee44d6eb4dec49bdfc9e81d schema:name Springer Nature - SN SciGraph project
    93 rdf:type schema:Organization
    94 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
    95 schema:name Information and Computing Sciences
    96 rdf:type schema:DefinedTerm
    97 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
    98 schema:name Artificial Intelligence and Image Processing
    99 rdf:type schema:DefinedTerm
    100 anzsrc-for:0806 schema:inDefinedTermSet anzsrc-for:
    101 schema:name Information Systems
    102 rdf:type schema:DefinedTerm
    103 anzsrc-for:17 schema:inDefinedTermSet anzsrc-for:
    104 schema:name Psychology and Cognitive Sciences
    105 rdf:type schema:DefinedTerm
    106 anzsrc-for:1702 schema:inDefinedTermSet anzsrc-for:
    107 schema:name Cognitive Sciences
    108 rdf:type schema:DefinedTerm
    109 sg:grant.2787267 http://pending.schema.org/fundedItem sg:pub.10.1007/s10994-013-5388-x
    110 rdf:type schema:MonetaryGrant
    111 sg:journal.1125588 schema:issn 0885-6125
    112 1573-0565
    113 schema:name Machine Learning
    114 schema:publisher Springer Nature
    115 rdf:type schema:Periodical
    116 sg:person.01206735310.03 schema:affiliation grid-institutes:grid.83440.3b
    117 schema:familyName Girolami
    118 schema:givenName M.
    119 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01206735310.03
    120 rdf:type schema:Person
    121 sg:person.012215654253.40 schema:affiliation grid-institutes:grid.30055.33
    122 schema:familyName Zhong
    123 schema:givenName M.
    124 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.012215654253.40
    125 rdf:type schema:Person
    126 sg:person.07706215665.03 schema:affiliation grid-institutes:grid.8756.c
    127 schema:familyName Filippone
    128 schema:givenName M.
    129 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07706215665.03
    130 rdf:type schema:Person
    131 sg:pub.10.1007/978-1-4612-0745-0 schema:sameAs https://app.dimensions.ai/details/publication/pub.1052635697
    132 https://doi.org/10.1007/978-1-4612-0745-0
    133 rdf:type schema:CreativeWork
    134 sg:pub.10.1023/a:1023562417138 schema:sameAs https://app.dimensions.ai/details/publication/pub.1036982513
    135 https://doi.org/10.1023/a:1023562417138
    136 rdf:type schema:CreativeWork
    137 grid-institutes:grid.30055.33 schema:alternateName Department of Biomedical Engineering, Dalian University of Technology, Dalian, P.R. China
    138 schema:name Department of Biomedical Engineering, Dalian University of Technology, Dalian, P.R. China
    139 rdf:type schema:Organization
    140 grid-institutes:grid.83440.3b schema:alternateName Department of Statistical Science, University College London, London, UK
    141 schema:name Department of Statistical Science, University College London, London, UK
    142 rdf:type schema:Organization
    143 grid-institutes:grid.8756.c schema:alternateName School of Computing Science, University of Glasgow, Glasgow, UK
    144 schema:name School of Computing Science, University of Glasgow, Glasgow, UK
    145 rdf:type schema:Organization
     




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


    ...