Probability Based Metrics for Nearest Neighbor Classification and Case-Based Reasoning View Full Text


Ontology type: schema:Chapter      Open Access: True


Chapter Info

DATE

1999-10-29

AUTHORS

Enrico Blanzieri , Francesco Ricci⋆

ABSTRACT

This paper is focused on a class of metrics for the Nearest Neighbor classifier, whose definition is based on statistics computed on the case base. We show that these metrics basically rely on a probability estimation phase. In particular, we reconsider a metric proposed in the 80’s by Short and Fukunaga, we extend its definition to an input space that includes categorical features and we evaluate empirically its performance. Moreover, we present an original probability based metric, called Minimum Risk Metric (MRM), i.e. a metric for classification tasks that exploits estimates of the posterior probabilities. MRM is optimal, in the sense that it optimizes the finite misclassification risk, whereas the Short and Fukunaga Metric minimize the difference between finite risk and asymptotic risk. An experimental comparison of MRM with the Short and Fukunaga Metric, the Value Difference Metric, and Euclidean-Hamming metrics on benchmark datasets shows that MRM outperforms the other metrics. MRM performs comparably to the Bayes Classifier based on the same probability estimates. The results suggest that MRM can be useful in case-based applications where the retrieval of a nearest neighbor is required. More... »

PAGES

14-28

Book

TITLE

Case-Based Reasoning Research and Development

ISBN

978-3-540-66237-2
978-3-540-48508-7

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/3-540-48508-2_2

DOI

http://dx.doi.org/10.1007/3-540-48508-2_2

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1004159122


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/2008/0104", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Statistics", 
        "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": {
          "name": [
            "Istituto per la Ricerca Scientifica e Tecnologica (ITC-IRST), 38050, Povo, TN, Italy"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Blanzieri", 
        "givenName": "Enrico", 
        "id": "sg:person.013033541655.32", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013033541655.32"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "name": [
            "Istituto per la Ricerca Scientifica e Tecnologica (ITC-IRST), 38050, Povo, TN, Italy"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Ricci\u22c6", 
        "givenName": "Francesco", 
        "id": "sg:person.014514650131.53", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014514650131.53"
        ], 
        "type": "Person"
      }
    ], 
    "citation": [
      {
        "id": "https://doi.org/10.1145/7902.7906", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1009203556"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/3-540-60598-3_27", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1020543938", 
          "https://doi.org/10.1007/3-540-60598-3_27"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/bf00993481", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1028790090", 
          "https://doi.org/10.1007/bf00993481"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1023/a:1007413511361", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1030336415", 
          "https://doi.org/10.1023/a:1007413511361"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/bf00994658", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1038601106", 
          "https://doi.org/10.1007/bf00994658"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1016/0031-3203(90)90123-3", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1043343713"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1016/0031-3203(90)90123-3", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1043343713"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "sg:pub.10.1007/bfb0056322", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1044029615", 
          "https://doi.org/10.1007/bfb0056322"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1145/135226.135228", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1044862785"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/34.761268", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061156947"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/tit.1967.1053964", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061646286"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1109/tit.1981.1056403", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1061648593"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1613/jair.346", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1105538442"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://doi.org/10.1002/9780470316849", 
        "sameAs": [
          "https://app.dimensions.ai/details/publication/pub.1109496273"
        ], 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://app.dimensions.ai/details/publication/pub.1109496273", 
        "type": "CreativeWork"
      }, 
      {
        "id": "https://app.dimensions.ai/details/publication/pub.1109496273", 
        "type": "CreativeWork"
      }
    ], 
    "datePublished": "1999-10-29", 
    "datePublishedReg": "1999-10-29", 
    "description": "This paper is focused on a class of metrics for the Nearest Neighbor classifier, whose definition is based on statistics computed on the case base. We show that these metrics basically rely on a probability estimation phase. In particular, we reconsider a metric proposed in the 80\u2019s by Short and Fukunaga, we extend its definition to an input space that includes categorical features and we evaluate empirically its performance. Moreover, we present an original probability based metric, called Minimum Risk Metric (MRM), i.e. a metric for classification tasks that exploits estimates of the posterior probabilities. MRM is optimal, in the sense that it optimizes the finite misclassification risk, whereas the Short and Fukunaga Metric minimize the difference between finite risk and asymptotic risk. An experimental comparison of MRM with the Short and Fukunaga Metric, the Value Difference Metric, and Euclidean-Hamming metrics on benchmark datasets shows that MRM outperforms the other metrics. MRM performs comparably to the Bayes Classifier based on the same probability estimates. The results suggest that MRM can be useful in case-based applications where the retrieval of a nearest neighbor is required.", 
    "editor": [
      {
        "familyName": "Althoff", 
        "givenName": "Klaus-Dieter", 
        "type": "Person"
      }, 
      {
        "familyName": "Bergmann", 
        "givenName": "Ralph", 
        "type": "Person"
      }, 
      {
        "familyName": "Branting", 
        "givenName": "L.Karl", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/3-540-48508-2_2", 
    "inLanguage": [
      "en"
    ], 
    "isAccessibleForFree": true, 
    "isPartOf": {
      "isbn": [
        "978-3-540-66237-2", 
        "978-3-540-48508-7"
      ], 
      "name": "Case-Based Reasoning Research and Development", 
      "type": "Book"
    }, 
    "name": "Probability Based Metrics for Nearest Neighbor Classification and Case-Based Reasoning", 
    "pagination": "14-28", 
    "productId": [
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/3-540-48508-2_2"
        ]
      }, 
      {
        "name": "readcube_id", 
        "type": "PropertyValue", 
        "value": [
          "13d1ae04353ae719a03bd95b6996e52934f6835698e1fd909c4c953a25719897"
        ]
      }, 
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1004159122"
        ]
      }
    ], 
    "publisher": {
      "location": "Berlin, Heidelberg", 
      "name": "Springer Berlin Heidelberg", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/3-540-48508-2_2", 
      "https://app.dimensions.ai/details/publication/pub.1004159122"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2019-04-16T05:37", 
    "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/v90/0000000346_0000000346/records_99836_00000000.jsonl", 
    "type": "Chapter", 
    "url": "https://link.springer.com/10.1007%2F3-540-48508-2_2"
  }
]
 

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.10.1007/3-540-48508-2_2'

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

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/3-540-48508-2_2'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/3-540-48508-2_2'

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

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/3-540-48508-2_2'


 

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

129 TRIPLES      23 PREDICATES      40 URIs      19 LITERALS      8 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/3-540-48508-2_2 schema:about anzsrc-for:01
2 anzsrc-for:0104
3 schema:author N3a9de110e7a54752b2a0c6c0d1080251
4 schema:citation sg:pub.10.1007/3-540-60598-3_27
5 sg:pub.10.1007/bf00993481
6 sg:pub.10.1007/bf00994658
7 sg:pub.10.1007/bfb0056322
8 sg:pub.10.1023/a:1007413511361
9 https://app.dimensions.ai/details/publication/pub.1109496273
10 https://doi.org/10.1002/9780470316849
11 https://doi.org/10.1016/0031-3203(90)90123-3
12 https://doi.org/10.1109/34.761268
13 https://doi.org/10.1109/tit.1967.1053964
14 https://doi.org/10.1109/tit.1981.1056403
15 https://doi.org/10.1145/135226.135228
16 https://doi.org/10.1145/7902.7906
17 https://doi.org/10.1613/jair.346
18 schema:datePublished 1999-10-29
19 schema:datePublishedReg 1999-10-29
20 schema:description This paper is focused on a class of metrics for the Nearest Neighbor classifier, whose definition is based on statistics computed on the case base. We show that these metrics basically rely on a probability estimation phase. In particular, we reconsider a metric proposed in the 80’s by Short and Fukunaga, we extend its definition to an input space that includes categorical features and we evaluate empirically its performance. Moreover, we present an original probability based metric, called Minimum Risk Metric (MRM), i.e. a metric for classification tasks that exploits estimates of the posterior probabilities. MRM is optimal, in the sense that it optimizes the finite misclassification risk, whereas the Short and Fukunaga Metric minimize the difference between finite risk and asymptotic risk. An experimental comparison of MRM with the Short and Fukunaga Metric, the Value Difference Metric, and Euclidean-Hamming metrics on benchmark datasets shows that MRM outperforms the other metrics. MRM performs comparably to the Bayes Classifier based on the same probability estimates. The results suggest that MRM can be useful in case-based applications where the retrieval of a nearest neighbor is required.
21 schema:editor N23b494945ca9428ca8722f8545218a4a
22 schema:genre chapter
23 schema:inLanguage en
24 schema:isAccessibleForFree true
25 schema:isPartOf N9e03119f39b34d39a245fc7c9f3756c1
26 schema:name Probability Based Metrics for Nearest Neighbor Classification and Case-Based Reasoning
27 schema:pagination 14-28
28 schema:productId N44de5f33ac9b4428b2e5def46e76be1b
29 N50cc6751c9e14054a7f7789bad833275
30 Nac61a6adbc874e12a13978fe1e604722
31 schema:publisher N286e469457c141dba871ae1cd98a0f33
32 schema:sameAs https://app.dimensions.ai/details/publication/pub.1004159122
33 https://doi.org/10.1007/3-540-48508-2_2
34 schema:sdDatePublished 2019-04-16T05:37
35 schema:sdLicense https://scigraph.springernature.com/explorer/license/
36 schema:sdPublisher Na30f4c5ff5914e80805b62323540767a
37 schema:url https://link.springer.com/10.1007%2F3-540-48508-2_2
38 sgo:license sg:explorer/license/
39 sgo:sdDataset chapters
40 rdf:type schema:Chapter
41 N0080e4d8a4404a949a21db335ac1fb4a schema:name Istituto per la Ricerca Scientifica e Tecnologica (ITC-IRST), 38050, Povo, TN, Italy
42 rdf:type schema:Organization
43 N0ad4f561f2b74a6f86dda19a50a85a83 rdf:first N4d2e1c85a33a4fc687b41a63aca796e5
44 rdf:rest Nde694e8ee3e24905b8db4fe07d535419
45 N23b494945ca9428ca8722f8545218a4a rdf:first N814b210f995540ea81240b475c4455ff
46 rdf:rest N0ad4f561f2b74a6f86dda19a50a85a83
47 N286e469457c141dba871ae1cd98a0f33 schema:location Berlin, Heidelberg
48 schema:name Springer Berlin Heidelberg
49 rdf:type schema:Organisation
50 N3a9de110e7a54752b2a0c6c0d1080251 rdf:first sg:person.013033541655.32
51 rdf:rest N4af2b02c323548bd811c4ac2e67ca606
52 N44de5f33ac9b4428b2e5def46e76be1b schema:name dimensions_id
53 schema:value pub.1004159122
54 rdf:type schema:PropertyValue
55 N4af2b02c323548bd811c4ac2e67ca606 rdf:first sg:person.014514650131.53
56 rdf:rest rdf:nil
57 N4d2e1c85a33a4fc687b41a63aca796e5 schema:familyName Bergmann
58 schema:givenName Ralph
59 rdf:type schema:Person
60 N50cc6751c9e14054a7f7789bad833275 schema:name doi
61 schema:value 10.1007/3-540-48508-2_2
62 rdf:type schema:PropertyValue
63 N62be8e247468462283bb87dabf7cbbbf schema:familyName Branting
64 schema:givenName L.Karl
65 rdf:type schema:Person
66 N814b210f995540ea81240b475c4455ff schema:familyName Althoff
67 schema:givenName Klaus-Dieter
68 rdf:type schema:Person
69 N9e03119f39b34d39a245fc7c9f3756c1 schema:isbn 978-3-540-48508-7
70 978-3-540-66237-2
71 schema:name Case-Based Reasoning Research and Development
72 rdf:type schema:Book
73 Na30f4c5ff5914e80805b62323540767a schema:name Springer Nature - SN SciGraph project
74 rdf:type schema:Organization
75 Nac61a6adbc874e12a13978fe1e604722 schema:name readcube_id
76 schema:value 13d1ae04353ae719a03bd95b6996e52934f6835698e1fd909c4c953a25719897
77 rdf:type schema:PropertyValue
78 Nde694e8ee3e24905b8db4fe07d535419 rdf:first N62be8e247468462283bb87dabf7cbbbf
79 rdf:rest rdf:nil
80 Nf91ee26ecb264ddea37d9cf95637ea61 schema:name Istituto per la Ricerca Scientifica e Tecnologica (ITC-IRST), 38050, Povo, TN, Italy
81 rdf:type schema:Organization
82 anzsrc-for:01 schema:inDefinedTermSet anzsrc-for:
83 schema:name Mathematical Sciences
84 rdf:type schema:DefinedTerm
85 anzsrc-for:0104 schema:inDefinedTermSet anzsrc-for:
86 schema:name Statistics
87 rdf:type schema:DefinedTerm
88 sg:person.013033541655.32 schema:affiliation N0080e4d8a4404a949a21db335ac1fb4a
89 schema:familyName Blanzieri
90 schema:givenName Enrico
91 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013033541655.32
92 rdf:type schema:Person
93 sg:person.014514650131.53 schema:affiliation Nf91ee26ecb264ddea37d9cf95637ea61
94 schema:familyName Ricci⋆
95 schema:givenName Francesco
96 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.014514650131.53
97 rdf:type schema:Person
98 sg:pub.10.1007/3-540-60598-3_27 schema:sameAs https://app.dimensions.ai/details/publication/pub.1020543938
99 https://doi.org/10.1007/3-540-60598-3_27
100 rdf:type schema:CreativeWork
101 sg:pub.10.1007/bf00993481 schema:sameAs https://app.dimensions.ai/details/publication/pub.1028790090
102 https://doi.org/10.1007/bf00993481
103 rdf:type schema:CreativeWork
104 sg:pub.10.1007/bf00994658 schema:sameAs https://app.dimensions.ai/details/publication/pub.1038601106
105 https://doi.org/10.1007/bf00994658
106 rdf:type schema:CreativeWork
107 sg:pub.10.1007/bfb0056322 schema:sameAs https://app.dimensions.ai/details/publication/pub.1044029615
108 https://doi.org/10.1007/bfb0056322
109 rdf:type schema:CreativeWork
110 sg:pub.10.1023/a:1007413511361 schema:sameAs https://app.dimensions.ai/details/publication/pub.1030336415
111 https://doi.org/10.1023/a:1007413511361
112 rdf:type schema:CreativeWork
113 https://app.dimensions.ai/details/publication/pub.1109496273 schema:CreativeWork
114 https://doi.org/10.1002/9780470316849 schema:sameAs https://app.dimensions.ai/details/publication/pub.1109496273
115 rdf:type schema:CreativeWork
116 https://doi.org/10.1016/0031-3203(90)90123-3 schema:sameAs https://app.dimensions.ai/details/publication/pub.1043343713
117 rdf:type schema:CreativeWork
118 https://doi.org/10.1109/34.761268 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061156947
119 rdf:type schema:CreativeWork
120 https://doi.org/10.1109/tit.1967.1053964 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061646286
121 rdf:type schema:CreativeWork
122 https://doi.org/10.1109/tit.1981.1056403 schema:sameAs https://app.dimensions.ai/details/publication/pub.1061648593
123 rdf:type schema:CreativeWork
124 https://doi.org/10.1145/135226.135228 schema:sameAs https://app.dimensions.ai/details/publication/pub.1044862785
125 rdf:type schema:CreativeWork
126 https://doi.org/10.1145/7902.7906 schema:sameAs https://app.dimensions.ai/details/publication/pub.1009203556
127 rdf:type schema:CreativeWork
128 https://doi.org/10.1613/jair.346 schema:sameAs https://app.dimensions.ai/details/publication/pub.1105538442
129 rdf:type schema:CreativeWork
 




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


...