A computationally efficient information estimator for weighted data

Hideitsu Hino; Noboru Murata

doi:10.1007/978-3-642-21738-8_39

A computationally efficient information estimator for weighted data

Hideitsu Hino^*, Noboru Murata

^*この研究の対応する著者

先進理工学部

研究成果: Conference contribution

3 被引用数 (Scopus)

抄録

The Shannon information content is a fundamental quantity and it is of great importance to estimate it from observed dataset in the field of statistics, information theory, and machine learning. In this study, an estimator for the information content using a given set of weighted data is proposed. The empirical data distribution varies depending on the weight. The notable features of the proposed estimator are its computational efficiency and its ability to deal with weighted data. The proposed estimator is extended in order to estimate cross entropy, entropy and KL divergence with weighted data. Then, the estimators are applied to classification with one-class samples, and distribution preserving data compression problems.

本文言語	English
ホスト出版物のタイトル	Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings
ページ	301-308
ページ数	8
版	PART 2
DOI	https://doi.org/10.1007/978-3-642-21738-8_39
出版ステータス	Published - 2011
イベント	21st International Conference on Artificial Neural Networks, ICANN 2011 - Espoo, Finland 継続期間: 2011 6月 14 → 2011 6月 17

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)
番号	PART 2
巻	6792 LNCS
ISSN（印刷版）	0302-9743
ISSN（電子版）	1611-3349

Conference

Conference	21st International Conference on Artificial Neural Networks, ICANN 2011
国/地域	Finland
City	Espoo
Period	11/6/14 → 11/6/17

ASJC Scopus subject areas

理論的コンピュータサイエンス
コンピュータサイエンス（全般）

文献へのアクセス

10.1007/978-3-642-21738-8_39

他のファイルとリンク

引用スタイル

Hino, H., & Murata, N. (2011). A computationally efficient information estimator for weighted data. In Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings (PART 2 ed., pp. 301-308). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6792 LNCS, No. PART 2). https://doi.org/10.1007/978-3-642-21738-8_39

A computationally efficient information estimator for weighted data. / Hino, Hideitsu; Murata, Noboru.
Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 2. ed. 2011. p. 301-308 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6792 LNCS, No. PART 2).

研究成果: Conference contribution

Hino, H & Murata, N 2011, A computationally efficient information estimator for weighted data. in Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 6792 LNCS, pp. 301-308, 21st International Conference on Artificial Neural Networks, ICANN 2011, Espoo, Finland, 11/6/14. https://doi.org/10.1007/978-3-642-21738-8_39

Hino H, Murata N. A computationally efficient information estimator for weighted data. In Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 2 ed. 2011. p. 301-308. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/978-3-642-21738-8_39

Hino, Hideitsu ; Murata, Noboru. / A computationally efficient information estimator for weighted data. Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 2. ed. 2011. pp. 301-308 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{87dd84bff8604eb2b689367c9c310ef3,

title = "A computationally efficient information estimator for weighted data",

abstract = "The Shannon information content is a fundamental quantity and it is of great importance to estimate it from observed dataset in the field of statistics, information theory, and machine learning. In this study, an estimator for the information content using a given set of weighted data is proposed. The empirical data distribution varies depending on the weight. The notable features of the proposed estimator are its computational efficiency and its ability to deal with weighted data. The proposed estimator is extended in order to estimate cross entropy, entropy and KL divergence with weighted data. Then, the estimators are applied to classification with one-class samples, and distribution preserving data compression problems.",

keywords = "Information, entropy, non-parametric, quantile",

author = "Hideitsu Hino and Noboru Murata",

year = "2011",

doi = "10.1007/978-3-642-21738-8_39",

language = "English",

isbn = "9783642217371",

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

number = "PART 2",

pages = "301--308",

booktitle = "Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings",

edition = "PART 2",

note = "21st International Conference on Artificial Neural Networks, ICANN 2011 ; Conference date: 14-06-2011 Through 17-06-2011",

}

TY - GEN

T1 - A computationally efficient information estimator for weighted data

AU - Hino, Hideitsu

AU - Murata, Noboru

PY - 2011

Y1 - 2011

N2 - The Shannon information content is a fundamental quantity and it is of great importance to estimate it from observed dataset in the field of statistics, information theory, and machine learning. In this study, an estimator for the information content using a given set of weighted data is proposed. The empirical data distribution varies depending on the weight. The notable features of the proposed estimator are its computational efficiency and its ability to deal with weighted data. The proposed estimator is extended in order to estimate cross entropy, entropy and KL divergence with weighted data. Then, the estimators are applied to classification with one-class samples, and distribution preserving data compression problems.

AB - The Shannon information content is a fundamental quantity and it is of great importance to estimate it from observed dataset in the field of statistics, information theory, and machine learning. In this study, an estimator for the information content using a given set of weighted data is proposed. The empirical data distribution varies depending on the weight. The notable features of the proposed estimator are its computational efficiency and its ability to deal with weighted data. The proposed estimator is extended in order to estimate cross entropy, entropy and KL divergence with weighted data. Then, the estimators are applied to classification with one-class samples, and distribution preserving data compression problems.

KW - Information

KW - entropy

KW - non-parametric

KW - quantile

UR - http://www.scopus.com/inward/record.url?scp=79959347694&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79959347694&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-21738-8_39

DO - 10.1007/978-3-642-21738-8_39

M3 - Conference contribution

AN - SCOPUS:79959347694

SN - 9783642217371

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

SP - 301

EP - 308

BT - Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings

T2 - 21st International Conference on Artificial Neural Networks, ICANN 2011

Y2 - 14 June 2011 through 17 June 2011

ER -

A computationally efficient information estimator for weighted data

抄録

出版物シリーズ

Conference

ASJC Scopus subject areas

文献へのアクセス

他のファイルとリンク

フィンガープリント

引用スタイル