Smooth boosting for margin-based ranking

研究成果: ジャーナルへの寄稿Conference article

3 引用 (Scopus)

抄録

We propose a new boosting algorithm for bipartite ranking problems. Our boosting algorithm, called SoftRankBoost, is a modification of RankBoost which maintains only smooth distributions over data. SoftRankBoost provably achieves approximately the maximum soft margin over all pairs of positive and negative examples, which implies high AUC score for future data.

元の言語英語
ページ(範囲)227-239
ページ数13
ジャーナルLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
5254 LNAI
DOI
出版物ステータス出版済み - 12 1 2008
イベント19th International Conference on Algorithmic Learning Theory, ALT 2008 - Budapest, ハンガリー
継続期間: 10 13 200810 16 2008

Fingerprint

Boosting
Margin
Ranking
Imply

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

これを引用

@article{18f55b20f5bd4bdc9b955e191648d0ca,
title = "Smooth boosting for margin-based ranking",
abstract = "We propose a new boosting algorithm for bipartite ranking problems. Our boosting algorithm, called SoftRankBoost, is a modification of RankBoost which maintains only smooth distributions over data. SoftRankBoost provably achieves approximately the maximum soft margin over all pairs of positive and negative examples, which implies high AUC score for future data.",
author = "Moribe, {Jun Ichi} and Kohei Hatano and Eiji Takimoto and Masayuki Takeda",
year = "2008",
month = "12",
day = "1",
doi = "10.1007/978-3-540-87987-9_21",
language = "English",
volume = "5254 LNAI",
pages = "227--239",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Smooth boosting for margin-based ranking

AU - Moribe, Jun Ichi

AU - Hatano, Kohei

AU - Takimoto, Eiji

AU - Takeda, Masayuki

PY - 2008/12/1

Y1 - 2008/12/1

N2 - We propose a new boosting algorithm for bipartite ranking problems. Our boosting algorithm, called SoftRankBoost, is a modification of RankBoost which maintains only smooth distributions over data. SoftRankBoost provably achieves approximately the maximum soft margin over all pairs of positive and negative examples, which implies high AUC score for future data.

AB - We propose a new boosting algorithm for bipartite ranking problems. Our boosting algorithm, called SoftRankBoost, is a modification of RankBoost which maintains only smooth distributions over data. SoftRankBoost provably achieves approximately the maximum soft margin over all pairs of positive and negative examples, which implies high AUC score for future data.

UR - http://www.scopus.com/inward/record.url?scp=56749160429&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=56749160429&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-87987-9_21

DO - 10.1007/978-3-540-87987-9_21

M3 - Conference article

AN - SCOPUS:56749160429

VL - 5254 LNAI

SP - 227

EP - 239

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -