On-line algorithm to predict nearly as well as the best pruning of a decision tree

Akira Maruoka, Eiji Takimoto

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

We review underlying mechanisms of the multiplicative weight-update prediction algorithms which somehow combine experts' predictions to obtain its own prediction that is almost as good as the best expert's prediction. Looking into the mechanisms we show how such an algorithm with the experts arranged on one layer can be naturally generalized to the one with the experts laid on nodes of trees. Consequently we give an on-line prediction algorithm that, when given a decision tree, produces predictions not much worse than the predictions made by the best pruning of the given decision tree.

Original languageEnglish
Title of host publicationProgress in Discovery Science
Pages296-306
Number of pages11
Volume2281
Publication statusPublished - 2002
Externally publishedYes

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2281
ISSN (Print)03029743
ISSN (Electronic)16113349

Fingerprint

Decision trees
Pruning
Decision tree
Predict
Prediction
Multiplicative
Update
Vertex of a graph

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Maruoka, A., & Takimoto, E. (2002). On-line algorithm to predict nearly as well as the best pruning of a decision tree. In Progress in Discovery Science (Vol. 2281, pp. 296-306). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2281).

On-line algorithm to predict nearly as well as the best pruning of a decision tree. / Maruoka, Akira; Takimoto, Eiji.

Progress in Discovery Science. Vol. 2281 2002. p. 296-306 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2281).

Research output: Chapter in Book/Report/Conference proceedingChapter

Maruoka, A & Takimoto, E 2002, On-line algorithm to predict nearly as well as the best pruning of a decision tree. in Progress in Discovery Science. vol. 2281, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2281, pp. 296-306.
Maruoka A, Takimoto E. On-line algorithm to predict nearly as well as the best pruning of a decision tree. In Progress in Discovery Science. Vol. 2281. 2002. p. 296-306. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Maruoka, Akira ; Takimoto, Eiji. / On-line algorithm to predict nearly as well as the best pruning of a decision tree. Progress in Discovery Science. Vol. 2281 2002. pp. 296-306 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inbook{267dc5e25ab049ea8f923d8a072625de,
title = "On-line algorithm to predict nearly as well as the best pruning of a decision tree",
abstract = "We review underlying mechanisms of the multiplicative weight-update prediction algorithms which somehow combine experts' predictions to obtain its own prediction that is almost as good as the best expert's prediction. Looking into the mechanisms we show how such an algorithm with the experts arranged on one layer can be naturally generalized to the one with the experts laid on nodes of trees. Consequently we give an on-line prediction algorithm that, when given a decision tree, produces predictions not much worse than the predictions made by the best pruning of the given decision tree.",
author = "Akira Maruoka and Eiji Takimoto",
year = "2002",
language = "English",
isbn = "3540433384",
volume = "2281",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "296--306",
booktitle = "Progress in Discovery Science",

}

TY - CHAP

T1 - On-line algorithm to predict nearly as well as the best pruning of a decision tree

AU - Maruoka, Akira

AU - Takimoto, Eiji

PY - 2002

Y1 - 2002

N2 - We review underlying mechanisms of the multiplicative weight-update prediction algorithms which somehow combine experts' predictions to obtain its own prediction that is almost as good as the best expert's prediction. Looking into the mechanisms we show how such an algorithm with the experts arranged on one layer can be naturally generalized to the one with the experts laid on nodes of trees. Consequently we give an on-line prediction algorithm that, when given a decision tree, produces predictions not much worse than the predictions made by the best pruning of the given decision tree.

AB - We review underlying mechanisms of the multiplicative weight-update prediction algorithms which somehow combine experts' predictions to obtain its own prediction that is almost as good as the best expert's prediction. Looking into the mechanisms we show how such an algorithm with the experts arranged on one layer can be naturally generalized to the one with the experts laid on nodes of trees. Consequently we give an on-line prediction algorithm that, when given a decision tree, produces predictions not much worse than the predictions made by the best pruning of the given decision tree.

UR - http://www.scopus.com/inward/record.url?scp=84867656064&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867656064&partnerID=8YFLogxK

M3 - Chapter

AN - SCOPUS:84867656064

SN - 3540433384

SN - 9783540433385

VL - 2281

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

SP - 296

EP - 306

BT - Progress in Discovery Science

ER -