Boosting over non-deterministic ZDDs

Takahiro Fujita, Kohei Hatano, Eiji Takimoto

研究成果: ジャーナルへの寄稿記事

抄録

We propose a new approach to large-scale machine learning, learning over compressed data: First compress the training data somehow and then employ various machine learning algorithms on the compressed data, with the hope that the computation time is significantly reduced when the training data is well compressed. As a first step toward this approach, we consider a variant of the Zero-Suppressed Binary Decision Diagram (ZDD) as the data structure for representing the training data, which is a generalization of the ZDD by incorporating non-determinism. For the learning algorithm to be employed, we consider a boosting algorithm called AdaBoost⁎ and its precursor AdaBoost. In this paper, we give efficient implementations of the boosting algorithms whose running times (per iteration) are linear in the size of the given ZDD.

元の言語英語
ページ(範囲)81-89
ページ数9
ジャーナルTheoretical Computer Science
806
DOI
出版物ステータス出版済み - 2 2 2020

Fingerprint

Binary decision diagrams
Boosting
Adaptive boosting
Decision Diagrams
Learning algorithms
Learning systems
AdaBoost
Binary
Learning Algorithm
Machine Learning
Zero
Data structures
Nondeterminism
Efficient Implementation
Precursor
Data Structures
Iteration
Training

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

これを引用

Boosting over non-deterministic ZDDs. / Fujita, Takahiro; Hatano, Kohei; Takimoto, Eiji.

:: Theoretical Computer Science, 巻 806, 02.02.2020, p. 81-89.

研究成果: ジャーナルへの寄稿記事

@article{4a92f28d33214eeaaf393056e0963d2e,
title = "Boosting over non-deterministic ZDDs",
abstract = "We propose a new approach to large-scale machine learning, learning over compressed data: First compress the training data somehow and then employ various machine learning algorithms on the compressed data, with the hope that the computation time is significantly reduced when the training data is well compressed. As a first step toward this approach, we consider a variant of the Zero-Suppressed Binary Decision Diagram (ZDD) as the data structure for representing the training data, which is a generalization of the ZDD by incorporating non-determinism. For the learning algorithm to be employed, we consider a boosting algorithm called AdaBoost⁎ and its precursor AdaBoost. In this paper, we give efficient implementations of the boosting algorithms whose running times (per iteration) are linear in the size of the given ZDD.",
author = "Takahiro Fujita and Kohei Hatano and Eiji Takimoto",
year = "2020",
month = "2",
day = "2",
doi = "10.1016/j.tcs.2018.11.027",
language = "English",
volume = "806",
pages = "81--89",
journal = "Theoretical Computer Science",
issn = "0304-3975",
publisher = "Elsevier",

}

TY - JOUR

T1 - Boosting over non-deterministic ZDDs

AU - Fujita, Takahiro

AU - Hatano, Kohei

AU - Takimoto, Eiji

PY - 2020/2/2

Y1 - 2020/2/2

N2 - We propose a new approach to large-scale machine learning, learning over compressed data: First compress the training data somehow and then employ various machine learning algorithms on the compressed data, with the hope that the computation time is significantly reduced when the training data is well compressed. As a first step toward this approach, we consider a variant of the Zero-Suppressed Binary Decision Diagram (ZDD) as the data structure for representing the training data, which is a generalization of the ZDD by incorporating non-determinism. For the learning algorithm to be employed, we consider a boosting algorithm called AdaBoost⁎ and its precursor AdaBoost. In this paper, we give efficient implementations of the boosting algorithms whose running times (per iteration) are linear in the size of the given ZDD.

AB - We propose a new approach to large-scale machine learning, learning over compressed data: First compress the training data somehow and then employ various machine learning algorithms on the compressed data, with the hope that the computation time is significantly reduced when the training data is well compressed. As a first step toward this approach, we consider a variant of the Zero-Suppressed Binary Decision Diagram (ZDD) as the data structure for representing the training data, which is a generalization of the ZDD by incorporating non-determinism. For the learning algorithm to be employed, we consider a boosting algorithm called AdaBoost⁎ and its precursor AdaBoost. In this paper, we give efficient implementations of the boosting algorithms whose running times (per iteration) are linear in the size of the given ZDD.

UR - http://www.scopus.com/inward/record.url?scp=85058165937&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85058165937&partnerID=8YFLogxK

U2 - 10.1016/j.tcs.2018.11.027

DO - 10.1016/j.tcs.2018.11.027

M3 - Article

AN - SCOPUS:85058165937

VL - 806

SP - 81

EP - 89

JO - Theoretical Computer Science

JF - Theoretical Computer Science

SN - 0304-3975

ER -