Essec\Faculty\Model\Contribution {#2216
  #_index: "academ_contributions"
  #_id: "10562"
  #_source: array:26 [
    "id" => "10562"
    "slug" => "probabilistic-low-rank-matrix-completion-on-finite-alphabets"
    "yearMonth" => "2014-12"
    "year" => "2014"
    "title" => "Probabilistic low-rank matrix completion on finite alphabets"
    "description" => "KLOPP, O., LAFOND, J., MOULINES, E. et SALMON, J. (2014). Probabilistic low-rank matrix completion on finite alphabets. Dans: <i>NIPS</i>. Montréal:  Neural Information Processing Systems."
    "authors" => array:4 [
      0 => array:3 [
        "name" => "KLOPP Olga"
        "bid" => "B00732676"
        "slug" => "klopp-olga"
      ]
      1 => array:1 [
        "name" => "LAFOND J."
      ]
      2 => array:1 [
        "name" => "MOULINES E."
      ]
      3 => array:1 [
        "name" => "SALMON J."
      ]
    ]
    "ouvrage" => "NIPS"
    "keywords" => []
    "updatedAt" => "2021-07-13 14:31:38"
    "publicationUrl" => "http://papers.nips.cc/paper/5358-probabilistic-low-rank-matrix-completion-on-finite-alphabets"
    "publicationInfo" => array:3 [
      "pages" => null
      "volume" => null
      "number" => null
    ]
    "type" => array:2 [
      "fr" => "Actes d'une conférence"
      "en" => "Conference Proceedings"
    ]
    "support_type" => array:2 [
      "fr" => "Editeur"
      "en" => "Publisher"
    ]
    "countries" => array:2 [
      "fr" => null
      "en" => null
    ]
    "abstract" => array:2 [
      "fr" => """
        The task of reconstructing a matrix given a sample of observed entries is known\n
        as the matrix completion problem. It arises in a wide range of problems, including\n
        recommender systems, collaborative filtering, dimensionality reduction,\n
        image processing, quantum physics or multi-class classification to name a few.\n
        Most works have focused on recovering an unknown real-valued low-rank matrix\n
        from randomly sub-sampling its entries. Here, we investigate the case where\n
        the observations take a finite number of values, corresponding for examples to\n
        ratings in recommender systems or labels in multi-class classification. We also\n
        consider a general sampling scheme (not necessarily uniform) over the matrix\n
        entries. The performance of a nuclear-norm penalized estimator is analyzed theoretically.\n
        More precisely, we derive bounds for the Kullback-Leibler divergence\n
        between the true and estimated distributions. In practice, we have also proposed\n
        an efficient algorithm based on lifted coordinate gradient descent in order to tackle\n
        potentially high dimensional settings.
        """
      "en" => """
        The task of reconstructing a matrix given a sample of observed entries is known\n
        as the matrix completion problem. It arises in a wide range of problems, including\n
        recommender systems, collaborative filtering, dimensionality reduction,\n
        image processing, quantum physics or multi-class classification to name a few.\n
        Most works have focused on recovering an unknown real-valued low-rank matrix\n
        from randomly sub-sampling its entries. Here, we investigate the case where\n
        the observations take a finite number of values, corresponding for examples to\n
        ratings in recommender systems or labels in multi-class classification. We also\n
        consider a general sampling scheme (not necessarily uniform) over the matrix\n
        entries. The performance of a nuclear-norm penalized estimator is analyzed theoretically.\n
        More precisely, we derive bounds for the Kullback-Leibler divergence\n
        between the true and estimated distributions. In practice, we have also proposed\n
        an efficient algorithm based on lifted coordinate gradient descent in order to tackle\n
        potentially high dimensional settings.
        """
    ]
    "authors_fields" => array:2 [
      "fr" => "Systèmes d'Information, Data Analytics et Opérations"
      "en" => "Information Systems, Data Analytics and Operations"
    ]
    "indexedAt" => "2024-11-21T10:21:50.000Z"
    "docTitle" => "Probabilistic low-rank matrix completion on finite alphabets"
    "docSurtitle" => "Conference Proceedings"
    "authorNames" => "<a href="/cv/klopp-olga">KLOPP Olga</a>, LAFOND J., MOULINES E., SALMON J."
    "docDescription" => "<span class="document-property-authors">KLOPP Olga, LAFOND J., MOULINES E., SALMON J.</span><br><span class="document-property-authors_fields">Information Systems, Data Analytics and Operations</span> | <span class="document-property-year">2014</span>"
    "keywordList" => ""
    "docPreview" => "<b>Probabilistic low-rank matrix completion on finite alphabets</b><br><span>2014-12 | Conference Proceedings </span>"
    "docType" => "research"
    "publicationLink" => "<a href="http://papers.nips.cc/paper/5358-probabilistic-low-rank-matrix-completion-on-finite-alphabets" target="_blank">Probabilistic low-rank matrix completion on finite alphabets</a>"
  ]
  +lang: "en"
  +"_type": "_doc"
  +"_score": 9.064929
  +"parent": null
}