Classifiers based on optimal decision rules

Talha M. Amin; Igor Chikalov; Mikhail Moshkov; Beata Zielosko

doi:10.3233/FI-2013-901

Classifiers based on optimal decision rules

Talha M. Amin, Igor Chikalov, Mikhail Moshkov, Beata Zielosko

Research output: Contribution to journal › Article › peer-review

2 Scopus citations

Abstract

Based on dynamic programming approach we design algorithms for sequential optimization of exact and approximate decision rules relative to the length and coverage [3, 4]. In this paper, we use optimal rules to construct classifiers, and study two questions: (i) which rules are better from the point of view of classification-exact or approximate; and (ii) which order of optimization gives better results of classifier work: length, length+coverage, coverage, or coverage+length. Experimental results show that, on average, classifiers based on exact rules are better than classifiers based on approximate rules, and sequential optimization (length+coverage or coverage+length) is better than the ordinary optimization (length or coverage).

Original language	English (US)
Pages (from-to)	151-160
Number of pages	10
Journal	Fundamenta Informaticae
Volume	127
Issue number	1-4
DOIs	https://doi.org/10.3233/FI-2013-901
State	Published - Nov 25 2013

Bibliographical note

KAUST Repository Item: Exported on 2020-10-01
Acknowledgements: This research was supported by King Abdullah University of Science and Technology in the frameworks of joint project with Nizhni Novgorod State University "Novel Algorithms in Machine Learning and Computer Vision, and Their High Performance Implementations", Russian Federal Program "Research and Development in Prioritized Directions of Scientific-Technological Complex of Russia in 2007-2013". The authors wish to express their gratitude to anonymous reviewers for useful comments.

ASJC Scopus subject areas

Computational Theory and Mathematics
Algebra and Number Theory
Theoretical Computer Science
Information Systems

Access to Document

10.3233/FI-2013-901

Cite this

@article{346c8d61f84e4dc58185064e15ffd2a5,

title = "Classifiers based on optimal decision rules",

abstract = "Based on dynamic programming approach we design algorithms for sequential optimization of exact and approximate decision rules relative to the length and coverage [3, 4]. In this paper, we use optimal rules to construct classifiers, and study two questions: (i) which rules are better from the point of view of classification-exact or approximate; and (ii) which order of optimization gives better results of classifier work: length, length+coverage, coverage, or coverage+length. Experimental results show that, on average, classifiers based on exact rules are better than classifiers based on approximate rules, and sequential optimization (length+coverage or coverage+length) is better than the ordinary optimization (length or coverage).",

author = "Amin, {Talha M.} and Igor Chikalov and Mikhail Moshkov and Beata Zielosko",

note = "KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: This research was supported by King Abdullah University of Science and Technology in the frameworks of joint project with Nizhni Novgorod State University {"}Novel Algorithms in Machine Learning and Computer Vision, and Their High Performance Implementations{"}, Russian Federal Program {"}Research and Development in Prioritized Directions of Scientific-Technological Complex of Russia in 2007-2013{"}. The authors wish to express their gratitude to anonymous reviewers for useful comments.",

year = "2013",

month = nov,

day = "25",

doi = "10.3233/FI-2013-901",

language = "English (US)",

volume = "127",

pages = "151--160",

journal = "Fundamenta Informaticae",

issn = "0169-2968",

publisher = "IOS Press BV",

number = "1-4",

}

TY - JOUR

T1 - Classifiers based on optimal decision rules

AU - Amin, Talha M.

AU - Chikalov, Igor

AU - Moshkov, Mikhail

AU - Zielosko, Beata

N1 - KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: This research was supported by King Abdullah University of Science and Technology in the frameworks of joint project with Nizhni Novgorod State University "Novel Algorithms in Machine Learning and Computer Vision, and Their High Performance Implementations", Russian Federal Program "Research and Development in Prioritized Directions of Scientific-Technological Complex of Russia in 2007-2013". The authors wish to express their gratitude to anonymous reviewers for useful comments.

PY - 2013/11/25

Y1 - 2013/11/25

N2 - Based on dynamic programming approach we design algorithms for sequential optimization of exact and approximate decision rules relative to the length and coverage [3, 4]. In this paper, we use optimal rules to construct classifiers, and study two questions: (i) which rules are better from the point of view of classification-exact or approximate; and (ii) which order of optimization gives better results of classifier work: length, length+coverage, coverage, or coverage+length. Experimental results show that, on average, classifiers based on exact rules are better than classifiers based on approximate rules, and sequential optimization (length+coverage or coverage+length) is better than the ordinary optimization (length or coverage).

AB - Based on dynamic programming approach we design algorithms for sequential optimization of exact and approximate decision rules relative to the length and coverage [3, 4]. In this paper, we use optimal rules to construct classifiers, and study two questions: (i) which rules are better from the point of view of classification-exact or approximate; and (ii) which order of optimization gives better results of classifier work: length, length+coverage, coverage, or coverage+length. Experimental results show that, on average, classifiers based on exact rules are better than classifiers based on approximate rules, and sequential optimization (length+coverage or coverage+length) is better than the ordinary optimization (length or coverage).

UR - http://hdl.handle.net/10754/563101

UR - https://www.medra.org/servlet/aliasResolver?alias=iospress&doi=10.3233/FI-2013-901

UR - http://www.scopus.com/inward/record.url?scp=84887980965&partnerID=8YFLogxK

U2 - 10.3233/FI-2013-901

DO - 10.3233/FI-2013-901

M3 - Article

SN - 0169-2968

VL - 127

SP - 151

EP - 160

JO - Fundamenta Informaticae

JF - Fundamenta Informaticae

IS - 1-4

ER -

Classifiers based on optimal decision rules

Abstract

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this