A View of the Em Algorithm that Justifies Incremental, Sparse, and other Variants View Full Text


Ontology type: schema:Chapter      Open Access: True


Chapter Info

DATE

1998

AUTHORS

Radford M. Neal , Geoffrey E. Hinton

ABSTRACT

The EM algorithm performs maximum likelihood estimation for data in which some variables are unobserved. We present a function that resembles negative free energy and show that the M step maximizes this function with respect to the model parameters and the E step maximizes it with respect to the distribution over the unobserved variables. From this perspective, it is easy to justify an incremental variant of the EM algorithm in which the distribution for only one of the unobserved variables is recalculated in each E step. This variant is shown empirically to give faster convergence in a mixture estimation problem. A variant of the algorithm that exploits sparse conditional distributions is also described, and a wide range of other variant algorithms are also seen to be possible. More... »

PAGES

355-368

Book

TITLE

Learning in Graphical Models

ISBN

978-94-010-6104-9
978-94-011-5014-9

Author Affiliations

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-94-011-5014-9_12

DOI

http://dx.doi.org/10.1007/978-94-011-5014-9_12

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1009494930


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 Toronto", 
          "id": "https://www.grid.ac/institutes/grid.17063.33", 
          "name": [
            "Dept. of Statistics and Dept. of Computer Science, University of Toronto, Toronto, Ontario, Canada"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Neal", 
        "givenName": "Radford M.", 
        "id": "sg:person.07677172755.27", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07677172755.27"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "University of Toronto", 
          "id": "https://www.grid.ac/institutes/grid.17063.33", 
          "name": [
            "Department of Computer Science, University of Toronto, Toronto, Ontario, Canada"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Hinton", 
        "givenName": "Geoffrey E.", 
        "id": "sg:person.0615147542.17", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0615147542.17"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "https://doi.org/10.1016/0167-7152(86)90016-7", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1016253042"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1016/0167-7152(86)90016-7", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1016253042"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1111/1467-9868.00082", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1021869359"
        ], 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "1998", 
    "datePublishedReg": "1998-01-01", 
    "description": "The EM algorithm performs maximum likelihood estimation for data in which some variables are unobserved. We present a function that resembles negative free energy and show that the M step maximizes this function with respect to the model parameters and the E step maximizes it with respect to the distribution over the unobserved variables. From this perspective, it is easy to justify an incremental variant of the EM algorithm in which the distribution for only one of the unobserved variables is recalculated in each E step. This variant is shown empirically to give faster convergence in a mixture estimation problem. A variant of the algorithm that exploits sparse conditional distributions is also described, and a wide range of other variant algorithms are also seen to be possible.", 
    "editor": [
      {
        "familyName": "Jordan", 
        "givenName": "Michael I.", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-94-011-5014-9_12", 
    "inLanguage": [
      "en"
    ], 
    "isAccessibleForFree": true, 
    "isPartOf": {
      "isbn": [
        "978-94-010-6104-9", 
        "978-94-011-5014-9"
      ], 
      "name": "Learning in Graphical Models", 
      "type": "Book"
    }, 
    "name": "A View of the Em Algorithm that Justifies Incremental, Sparse, and other Variants", 
    "pagination": "355-368", 
    "productId": [
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-94-011-5014-9_12"
        ]
      }, 
      {
        "name": "readcube_id", 
        "type": "PropertyValue", 
        "value": [
          "2d3abc1f5dad12b0910bcec46e23443c19b1f7e08be428998e4ba18d9c5a2b89"
        ]
      }, 
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1009494930"
        ]
      }
    ], 
    "publisher": {
      "location": "Dordrecht", 
      "name": "Springer Netherlands", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-94-011-5014-9_12", 
      "https://app.dimensions.ai/details/publication/pub.1009494930"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2019-04-15T17:47", 
    "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_8678_00000553.jsonl", 
    "type": "Chapter", 
    "url": "http://link.springer.com/10.1007/978-94-011-5014-9_12"
  }
]
 

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.10.1007/978-94-011-5014-9_12'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-94-011-5014-9_12'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-94-011-5014-9_12'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-94-011-5014-9_12'


 

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

79 TRIPLES      23 PREDICATES      29 URIs      20 LITERALS      8 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-94-011-5014-9_12 schema:about anzsrc-for:01
2 anzsrc-for:0103
3 schema:author N92b27e7e1759419ebfded07a3cc09170
4 schema:citation https://doi.org/10.1016/0167-7152(86)90016-7
5 https://doi.org/10.1111/1467-9868.00082
6 schema:datePublished 1998
7 schema:datePublishedReg 1998-01-01
8 schema:description The EM algorithm performs maximum likelihood estimation for data in which some variables are unobserved. We present a function that resembles negative free energy and show that the M step maximizes this function with respect to the model parameters and the E step maximizes it with respect to the distribution over the unobserved variables. From this perspective, it is easy to justify an incremental variant of the EM algorithm in which the distribution for only one of the unobserved variables is recalculated in each E step. This variant is shown empirically to give faster convergence in a mixture estimation problem. A variant of the algorithm that exploits sparse conditional distributions is also described, and a wide range of other variant algorithms are also seen to be possible.
9 schema:editor N1cc88f162d7b4106b3a786b6ee3f8feb
10 schema:genre chapter
11 schema:inLanguage en
12 schema:isAccessibleForFree true
13 schema:isPartOf Nbac5342e436f483aaed8289c44bb4837
14 schema:name A View of the Em Algorithm that Justifies Incremental, Sparse, and other Variants
15 schema:pagination 355-368
16 schema:productId Na6a62840c15a4f51908d454b7bdd7b09
17 Nd1f46882ded149f3a0e3ed13dca9e624
18 Nd23a1e6d6f7c45859c64f248f5d405c7
19 schema:publisher N03ee4e5221694667aa1667d9bd082a22
20 schema:sameAs https://app.dimensions.ai/details/publication/pub.1009494930
21 https://doi.org/10.1007/978-94-011-5014-9_12
22 schema:sdDatePublished 2019-04-15T17:47
23 schema:sdLicense https://scigraph.springernature.com/explorer/license/
24 schema:sdPublisher N1edcc4ac49fb4e7ab18ad6a3095b16e8
25 schema:url http://link.springer.com/10.1007/978-94-011-5014-9_12
26 sgo:license sg:explorer/license/
27 sgo:sdDataset chapters
28 rdf:type schema:Chapter
29 N03ee4e5221694667aa1667d9bd082a22 schema:location Dordrecht
30 schema:name Springer Netherlands
31 rdf:type schema:Organisation
32 N1cc88f162d7b4106b3a786b6ee3f8feb rdf:first N4a949867367c412ea0de06984d43982d
33 rdf:rest rdf:nil
34 N1edcc4ac49fb4e7ab18ad6a3095b16e8 schema:name Springer Nature - SN SciGraph project
35 rdf:type schema:Organization
36 N4a949867367c412ea0de06984d43982d schema:familyName Jordan
37 schema:givenName Michael I.
38 rdf:type schema:Person
39 N76b951b5fecc4d8c8504988369bb9f5e rdf:first sg:person.0615147542.17
40 rdf:rest rdf:nil
41 N92b27e7e1759419ebfded07a3cc09170 rdf:first sg:person.07677172755.27
42 rdf:rest N76b951b5fecc4d8c8504988369bb9f5e
43 Na6a62840c15a4f51908d454b7bdd7b09 schema:name readcube_id
44 schema:value 2d3abc1f5dad12b0910bcec46e23443c19b1f7e08be428998e4ba18d9c5a2b89
45 rdf:type schema:PropertyValue
46 Nbac5342e436f483aaed8289c44bb4837 schema:isbn 978-94-010-6104-9
47 978-94-011-5014-9
48 schema:name Learning in Graphical Models
49 rdf:type schema:Book
50 Nd1f46882ded149f3a0e3ed13dca9e624 schema:name doi
51 schema:value 10.1007/978-94-011-5014-9_12
52 rdf:type schema:PropertyValue
53 Nd23a1e6d6f7c45859c64f248f5d405c7 schema:name dimensions_id
54 schema:value pub.1009494930
55 rdf:type schema:PropertyValue
56 anzsrc-for:01 schema:inDefinedTermSet anzsrc-for:
57 schema:name Mathematical Sciences
58 rdf:type schema:DefinedTerm
59 anzsrc-for:0103 schema:inDefinedTermSet anzsrc-for:
60 schema:name Numerical and Computational Mathematics
61 rdf:type schema:DefinedTerm
62 sg:person.0615147542.17 schema:affiliation https://www.grid.ac/institutes/grid.17063.33
63 schema:familyName Hinton
64 schema:givenName Geoffrey E.
65 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0615147542.17
66 rdf:type schema:Person
67 sg:person.07677172755.27 schema:affiliation https://www.grid.ac/institutes/grid.17063.33
68 schema:familyName Neal
69 schema:givenName Radford M.
70 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.07677172755.27
71 rdf:type schema:Person
72 https://doi.org/10.1016/0167-7152(86)90016-7 schema:sameAs https://app.dimensions.ai/details/publication/pub.1016253042
73 rdf:type schema:CreativeWork
74 https://doi.org/10.1111/1467-9868.00082 schema:sameAs https://app.dimensions.ai/details/publication/pub.1021869359
75 rdf:type schema:CreativeWork
76 https://www.grid.ac/institutes/grid.17063.33 schema:alternateName University of Toronto
77 schema:name Department of Computer Science, University of Toronto, Toronto, Ontario, Canada
78 Dept. of Statistics and Dept. of Computer Science, University of Toronto, Toronto, Ontario, Canada
79 rdf:type schema:Organization
 




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


...