Optimization of approximate decision rules relative to number of misclassifications

Handle URI:
http://hdl.handle.net/10754/562452
Title:
Optimization of approximate decision rules relative to number of misclassifications
Authors:
Amin, Talha ( 0000-0003-3035-8612 ) ; Chikalov, Igor; Moshkov, Mikhail ( 0000-0003-0085-9483 ) ; Zielosko, Beata
Abstract:
In the paper, we study an extension of dynamic programming approach which allows optimization of approximate decision rules relative to the number of misclassifications. We introduce an uncertainty measure J(T) which is a difference between the number of rows in a decision table T and the number of rows with the most common decision for T. For a nonnegative real number γ, we consider γ-decision rules that localize rows in subtables of T with uncertainty at most γ. The presented algorithm constructs a directed acyclic graph Δγ(T). Based on this graph we can describe the whole set of so-called irredundant γ-decision rules. We can optimize rules from this set according to the number of misclassifications. Results of experiments with decision tables from the UCI Machine Learning Repository are presented. © 2012 The authors and IOS Press. All rights reserved.
KAUST Department:
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division; Applied Mathematics and Computational Science Program; Extensions of Dynamic Programming, Machine Learning and Discrete Optimization Research Group
Publisher:
IOS Press
Journal:
Frontiers in Artificial Intelligence and Applications
Issue Date:
1-Dec-2012
DOI:
10.3233/978-1-61499-105-2-674
Type:
Article
ISSN:
09226389
ISBN:
9781614991045
Appears in Collections:
Articles; Applied Mathematics and Computational Science Program; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorAmin, Talhaen
dc.contributor.authorChikalov, Igoren
dc.contributor.authorMoshkov, Mikhailen
dc.contributor.authorZielosko, Beataen
dc.date.accessioned2015-08-03T10:38:42Zen
dc.date.available2015-08-03T10:38:42Zen
dc.date.issued2012-12-01en
dc.identifier.isbn9781614991045en
dc.identifier.issn09226389en
dc.identifier.doi10.3233/978-1-61499-105-2-674en
dc.identifier.urihttp://hdl.handle.net/10754/562452en
dc.description.abstractIn the paper, we study an extension of dynamic programming approach which allows optimization of approximate decision rules relative to the number of misclassifications. We introduce an uncertainty measure J(T) which is a difference between the number of rows in a decision table T and the number of rows with the most common decision for T. For a nonnegative real number γ, we consider γ-decision rules that localize rows in subtables of T with uncertainty at most γ. The presented algorithm constructs a directed acyclic graph Δγ(T). Based on this graph we can describe the whole set of so-called irredundant γ-decision rules. We can optimize rules from this set according to the number of misclassifications. Results of experiments with decision tables from the UCI Machine Learning Repository are presented. © 2012 The authors and IOS Press. All rights reserved.en
dc.publisherIOS Pressen
dc.subjectdecision rulesen
dc.subjectdynamic programmingen
dc.subjectnumber of misclassificationsen
dc.titleOptimization of approximate decision rules relative to number of misclassificationsen
dc.typeArticleen
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.contributor.departmentApplied Mathematics and Computational Science Programen
dc.contributor.departmentExtensions of Dynamic Programming, Machine Learning and Discrete Optimization Research Groupen
dc.identifier.journalFrontiers in Artificial Intelligence and Applicationsen
dc.contributor.institutionInstitute of Computer Science, University of Silesia, 39, Bȩdziñska St., 41-200 Sosnowiec, Polanden
kaust.authorChikalov, Igoren
kaust.authorMoshkov, Mikhailen
kaust.authorZielosko, Beataen
kaust.authorAmin, Talhaen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.