Negative encoding length as a subjective interestingness measure for groups of rules

Einoshin Suzuki

doi:10.1007/978-3-642-01307-2_22

Negative encoding length as a subjective interestingness measure for groups of rules

Einoshin Suzuki

知能科学

研究成果: Chapter in Book/Report/Conference proceeding › Conference contribution

4 被引用数 (Scopus)

抄録

We propose an interestingness measure for groups of classification rules which are mutually related based on the Minimum Description Length Principle. Unlike conventional methods, our interestingness measure is based on a theoretical background, has no parameter, is applicable to a group of any number of rules, and can exploit an initial hypothesis. We have integrated the interestingness measure with practical heuristic search and built a rule-group discovery method CLARDEM (Classification Rule Discovery method based on an Extended-Mdlp).Extensive experiments using both real and artificial data confirm that CLARDEM can discover the correct concept from a small noisy data set and an approximate initial concept with high "discovery accuracy".

本文言語	英語
ホスト出版物のタイトル	13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009
ページ	220-231
ページ数	12
DOI	https://doi.org/10.1007/978-3-642-01307-2_22
出版ステータス	出版済み - 7 23 2009
イベント	13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009 - Bangkok, タイ継続期間: 4 27 2009 → 4 30 2009

出版物シリーズ

名前	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
巻	5476 LNAI
ISSN（印刷版）	0302-9743
ISSN（電子版）	1611-3349

その他

その他	13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009
Country	タイ
City	Bangkok
Period	4/27/09 → 4/30/09

All Science Journal Classification (ASJC) codes

Theoretical Computer Science
Computer Science(all)

文献へのアクセス

10.1007/978-3-642-01307-2_22

フィンガープリント「Negative encoding length as a subjective interestingness measure for groups of rules」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル

Suzuki, E. (2009). Negative encoding length as a subjective interestingness measure for groups of rules. In 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009 (pp. 220-231). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5476 LNAI). https://doi.org/10.1007/978-3-642-01307-2_22

Negative encoding length as a subjective interestingness measure for groups of rules. / Suzuki, Einoshin.

13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009. 2009. p. 220-231 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5476 LNAI).

研究成果: Chapter in Book/Report/Conference proceeding › Conference contribution

Suzuki, E 2009, Negative encoding length as a subjective interestingness measure for groups of rules. in 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5476 LNAI, pp. 220-231, 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009, Bangkok, タイ, 4/27/09. https://doi.org/10.1007/978-3-642-01307-2_22

Suzuki E. Negative encoding length as a subjective interestingness measure for groups of rules. In 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009. 2009. p. 220-231. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-01307-2_22

@inproceedings{7d3e38fad15b4821a16a49bad0d948c2,

title = "Negative encoding length as a subjective interestingness measure for groups of rules",

abstract = "We propose an interestingness measure for groups of classification rules which are mutually related based on the Minimum Description Length Principle. Unlike conventional methods, our interestingness measure is based on a theoretical background, has no parameter, is applicable to a group of any number of rules, and can exploit an initial hypothesis. We have integrated the interestingness measure with practical heuristic search and built a rule-group discovery method CLARDEM (Classification Rule Discovery method based on an Extended-Mdlp).Extensive experiments using both real and artificial data confirm that CLARDEM can discover the correct concept from a small noisy data set and an approximate initial concept with high {"}discovery accuracy{"}.",

author = "Einoshin Suzuki",

year = "2009",

month = jul,

day = "23",

doi = "10.1007/978-3-642-01307-2_22",

language = "English",

isbn = "3642013066",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "220--231",

booktitle = "13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009",

note = "13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009 ; Conference date: 27-04-2009 Through 30-04-2009",

}

TY - GEN

T1 - Negative encoding length as a subjective interestingness measure for groups of rules

AU - Suzuki, Einoshin

PY - 2009/7/23

Y1 - 2009/7/23

N2 - We propose an interestingness measure for groups of classification rules which are mutually related based on the Minimum Description Length Principle. Unlike conventional methods, our interestingness measure is based on a theoretical background, has no parameter, is applicable to a group of any number of rules, and can exploit an initial hypothesis. We have integrated the interestingness measure with practical heuristic search and built a rule-group discovery method CLARDEM (Classification Rule Discovery method based on an Extended-Mdlp).Extensive experiments using both real and artificial data confirm that CLARDEM can discover the correct concept from a small noisy data set and an approximate initial concept with high "discovery accuracy".

AB - We propose an interestingness measure for groups of classification rules which are mutually related based on the Minimum Description Length Principle. Unlike conventional methods, our interestingness measure is based on a theoretical background, has no parameter, is applicable to a group of any number of rules, and can exploit an initial hypothesis. We have integrated the interestingness measure with practical heuristic search and built a rule-group discovery method CLARDEM (Classification Rule Discovery method based on an Extended-Mdlp).Extensive experiments using both real and artificial data confirm that CLARDEM can discover the correct concept from a small noisy data set and an approximate initial concept with high "discovery accuracy".

UR - http://www.scopus.com/inward/record.url?scp=67650686502&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650686502&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-01307-2_22

DO - 10.1007/978-3-642-01307-2_22

M3 - Conference contribution

AN - SCOPUS:67650686502

SN - 3642013066

SN - 9783642013065

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 220

EP - 231

BT - 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009

T2 - 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2009

Y2 - 27 April 2009 through 30 April 2009

ER -