Pick Your Neighborhood – Improving Labels and Neighborhood Structure for Label Propagation View Full Text


Ontology type: schema:Chapter     


Chapter Info

DATE

2011

AUTHORS

Sandra Ebert , Mario Fritz , Bernt Schiele

ABSTRACT

Graph-based methods are very popular in semi-supervised learning due to their well founded theoretical background, intuitive interpretation of local neighborhood structure, and strong performance on a wide range of challenging learning problems. However, the success of these methods is highly dependent on the pre-existing neighborhood structure in the data used to construct the graph. In this paper, we use metric learning to improve this critical step by increasing the precision of the nearest neighbors and building our graph in this new metric space. We show that learning of neighborhood relations before constructing the graph consistently improves performance of two label propagation schemes on three different datasets – achieving the best performance reported on Caltech 101 to date. Furthermore, we question the predominant random draw of labels and advocate the importance of the choice of labeled examples. Orthogonal to active learning schemes, we investigate how domain knowledge can substantially increase performance in these semi-supervised learning settings. More... »

PAGES

152-162

Book

TITLE

Pattern Recognition

ISBN

978-3-642-23122-3
978-3-642-23123-0

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-642-23123-0_16

DOI

http://dx.doi.org/10.1007/978-3-642-23123-0_16

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1008140842


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/2008/13", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Education", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/1303", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Specialist Studies In Education", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "TU Darmstadt, Germany", 
          "id": "http://www.grid.ac/institutes/grid.6546.1", 
          "name": [
            "MPI Informatics, Saarbrucken, Germany", 
            "TU Darmstadt, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Ebert", 
        "givenName": "Sandra", 
        "id": "sg:person.011333635343.49", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.011333635343.49"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "MPI Informatics, Saarbrucken, Germany", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "MPI Informatics, Saarbrucken, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Fritz", 
        "givenName": "Mario", 
        "id": "sg:person.013361072755.17", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013361072755.17"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "MPI Informatics, Saarbrucken, Germany", 
          "id": "http://www.grid.ac/institutes/None", 
          "name": [
            "MPI Informatics, Saarbrucken, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Schiele", 
        "givenName": "Bernt", 
        "id": "sg:person.01174260421.90", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01174260421.90"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2011", 
    "datePublishedReg": "2011-01-01", 
    "description": "Graph-based methods are very popular in semi-supervised learning due to their well founded theoretical background, intuitive interpretation of local neighborhood structure, and strong performance on a wide range of challenging learning problems. However, the success of these methods is highly dependent on the pre-existing neighborhood structure in the data used to construct the graph. In this paper, we use metric learning to improve this critical step by increasing the precision of the nearest neighbors and building our graph in this new metric space. We show that learning of neighborhood relations before constructing the graph consistently improves performance of two label propagation schemes on three different datasets \u2013 achieving the best performance reported on Caltech 101 to date. Furthermore, we question the predominant random draw of labels and advocate the importance of the choice of labeled examples. Orthogonal to active learning schemes, we investigate how domain knowledge can substantially increase performance in these semi-supervised learning settings.", 
    "editor": [
      {
        "familyName": "Mester", 
        "givenName": "Rudolf", 
        "type": "Person"
      }, 
      {
        "familyName": "Felsberg", 
        "givenName": "Michael", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-642-23123-0_16", 
    "isAccessibleForFree": false, 
    "isPartOf": {
      "isbn": [
        "978-3-642-23122-3", 
        "978-3-642-23123-0"
      ], 
      "name": "Pattern Recognition", 
      "type": "Book"
    }, 
    "keywords": [
      "neighborhood structure", 
      "label propagation scheme", 
      "active learning scheme", 
      "semi-supervised learning", 
      "local neighborhood structure", 
      "Caltech-101", 
      "domain knowledge", 
      "learning problem", 
      "metric learning", 
      "learning scheme", 
      "label propagation", 
      "nearest neighbors", 
      "new metric space", 
      "strong performance", 
      "neighborhood relations", 
      "propagation scheme", 
      "graph", 
      "learning", 
      "better performance", 
      "intuitive interpretation", 
      "scheme", 
      "performance", 
      "labels", 
      "theoretical background", 
      "neighbors", 
      "metric spaces", 
      "critical step", 
      "method", 
      "wide range", 
      "precision", 
      "space", 
      "example", 
      "knowledge", 
      "data", 
      "step", 
      "structure", 
      "success", 
      "random draws", 
      "setting", 
      "choice", 
      "propagation", 
      "background", 
      "interpretation", 
      "importance", 
      "draw", 
      "range", 
      "relation", 
      "date", 
      "wells", 
      "problem", 
      "paper"
    ], 
    "name": "Pick Your Neighborhood \u2013 Improving Labels and Neighborhood Structure for Label Propagation", 
    "pagination": "152-162", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1008140842"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-642-23123-0_16"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-642-23123-0_16", 
      "https://app.dimensions.ai/details/publication/pub.1008140842"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-11-24T21:13", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20221124/entities/gbq_results/chapter/chapter_185.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-642-23123-0_16"
  }
]
 

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.10.1007/978-3-642-23123-0_16'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-23123-0_16'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-23123-0_16'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-642-23123-0_16'


 

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

133 TRIPLES      22 PREDICATES      76 URIs      69 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-642-23123-0_16 schema:about anzsrc-for:13
2 anzsrc-for:1303
3 schema:author N5a404d1265ca4aa182fb456f91501b1b
4 schema:datePublished 2011
5 schema:datePublishedReg 2011-01-01
6 schema:description Graph-based methods are very popular in semi-supervised learning due to their well founded theoretical background, intuitive interpretation of local neighborhood structure, and strong performance on a wide range of challenging learning problems. However, the success of these methods is highly dependent on the pre-existing neighborhood structure in the data used to construct the graph. In this paper, we use metric learning to improve this critical step by increasing the precision of the nearest neighbors and building our graph in this new metric space. We show that learning of neighborhood relations before constructing the graph consistently improves performance of two label propagation schemes on three different datasets – achieving the best performance reported on Caltech 101 to date. Furthermore, we question the predominant random draw of labels and advocate the importance of the choice of labeled examples. Orthogonal to active learning schemes, we investigate how domain knowledge can substantially increase performance in these semi-supervised learning settings.
7 schema:editor N2b061dc53db54d2ba72ba433ce7cd31a
8 schema:genre chapter
9 schema:isAccessibleForFree false
10 schema:isPartOf N2157b9a3bc184ce19e051e1a27e7cad7
11 schema:keywords Caltech-101
12 active learning scheme
13 background
14 better performance
15 choice
16 critical step
17 data
18 date
19 domain knowledge
20 draw
21 example
22 graph
23 importance
24 interpretation
25 intuitive interpretation
26 knowledge
27 label propagation
28 label propagation scheme
29 labels
30 learning
31 learning problem
32 learning scheme
33 local neighborhood structure
34 method
35 metric learning
36 metric spaces
37 nearest neighbors
38 neighborhood relations
39 neighborhood structure
40 neighbors
41 new metric space
42 paper
43 performance
44 precision
45 problem
46 propagation
47 propagation scheme
48 random draws
49 range
50 relation
51 scheme
52 semi-supervised learning
53 setting
54 space
55 step
56 strong performance
57 structure
58 success
59 theoretical background
60 wells
61 wide range
62 schema:name Pick Your Neighborhood – Improving Labels and Neighborhood Structure for Label Propagation
63 schema:pagination 152-162
64 schema:productId N05ec8df184464ae0a8ad1993fb44b1ee
65 N1b0139d4fdf54abab31ebf2d1b646925
66 schema:publisher N1b65a07cf68c44168e73e11efd77a321
67 schema:sameAs https://app.dimensions.ai/details/publication/pub.1008140842
68 https://doi.org/10.1007/978-3-642-23123-0_16
69 schema:sdDatePublished 2022-11-24T21:13
70 schema:sdLicense https://scigraph.springernature.com/explorer/license/
71 schema:sdPublisher N4f50515bab7b4c81802b631ba8ec0ecf
72 schema:url https://doi.org/10.1007/978-3-642-23123-0_16
73 sgo:license sg:explorer/license/
74 sgo:sdDataset chapters
75 rdf:type schema:Chapter
76 N05ec8df184464ae0a8ad1993fb44b1ee schema:name dimensions_id
77 schema:value pub.1008140842
78 rdf:type schema:PropertyValue
79 N1b0139d4fdf54abab31ebf2d1b646925 schema:name doi
80 schema:value 10.1007/978-3-642-23123-0_16
81 rdf:type schema:PropertyValue
82 N1b65a07cf68c44168e73e11efd77a321 schema:name Springer Nature
83 rdf:type schema:Organisation
84 N20e3dc0548044aae80b85da275de33ac rdf:first sg:person.013361072755.17
85 rdf:rest Nd23c1d6fe023413eb878c0dd28d179f9
86 N212ec5a5ad52431caca92e918dd28386 schema:familyName Felsberg
87 schema:givenName Michael
88 rdf:type schema:Person
89 N2157b9a3bc184ce19e051e1a27e7cad7 schema:isbn 978-3-642-23122-3
90 978-3-642-23123-0
91 schema:name Pattern Recognition
92 rdf:type schema:Book
93 N2b061dc53db54d2ba72ba433ce7cd31a rdf:first Nd10e9e28f0444454bd1d2b684d376279
94 rdf:rest Neea511824c3843dba69595c5143240fe
95 N4f50515bab7b4c81802b631ba8ec0ecf schema:name Springer Nature - SN SciGraph project
96 rdf:type schema:Organization
97 N5a404d1265ca4aa182fb456f91501b1b rdf:first sg:person.011333635343.49
98 rdf:rest N20e3dc0548044aae80b85da275de33ac
99 Nd10e9e28f0444454bd1d2b684d376279 schema:familyName Mester
100 schema:givenName Rudolf
101 rdf:type schema:Person
102 Nd23c1d6fe023413eb878c0dd28d179f9 rdf:first sg:person.01174260421.90
103 rdf:rest rdf:nil
104 Neea511824c3843dba69595c5143240fe rdf:first N212ec5a5ad52431caca92e918dd28386
105 rdf:rest rdf:nil
106 anzsrc-for:13 schema:inDefinedTermSet anzsrc-for:
107 schema:name Education
108 rdf:type schema:DefinedTerm
109 anzsrc-for:1303 schema:inDefinedTermSet anzsrc-for:
110 schema:name Specialist Studies In Education
111 rdf:type schema:DefinedTerm
112 sg:person.011333635343.49 schema:affiliation grid-institutes:grid.6546.1
113 schema:familyName Ebert
114 schema:givenName Sandra
115 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.011333635343.49
116 rdf:type schema:Person
117 sg:person.01174260421.90 schema:affiliation grid-institutes:None
118 schema:familyName Schiele
119 schema:givenName Bernt
120 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01174260421.90
121 rdf:type schema:Person
122 sg:person.013361072755.17 schema:affiliation grid-institutes:None
123 schema:familyName Fritz
124 schema:givenName Mario
125 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013361072755.17
126 rdf:type schema:Person
127 grid-institutes:None schema:alternateName MPI Informatics, Saarbrucken, Germany
128 schema:name MPI Informatics, Saarbrucken, Germany
129 rdf:type schema:Organization
130 grid-institutes:grid.6546.1 schema:alternateName TU Darmstadt, Germany
131 schema:name MPI Informatics, Saarbrucken, Germany
132 TU Darmstadt, Germany
133 rdf:type schema:Organization
 




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


...