Total Path Length and Number of Terminal Nodes for Decision Trees

Handle URI:
http://hdl.handle.net/10754/552376
Title:
Total Path Length and Number of Terminal Nodes for Decision Trees
Authors:
Hussain, Shahid ( 0000-0002-1698-2809 )
Abstract:
This paper presents a new tool for study of relationships between total path length (average depth) and number of terminal nodes for decision trees. These relationships are important from the point of view of optimization of decision trees. In this particular case of total path length and number of terminal nodes, the relationships between these two cost functions are closely related with space-time trade-off. In addition to algorithm to compute the relationships, the paper also presents results of experiments with datasets from UCI ML Repository1. These experiments show how two cost functions behave for a given decision table and the resulting plots show the Pareto frontier or Pareto set of optimal points. Furthermore, in some cases this Pareto frontier is a singleton showing the total optimality of decision trees for the given decision table.
KAUST Department:
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Citation:
Total Path Length and Number of Terminal Nodes for Decision Trees 2014, 35:514 Procedia Computer Science
Publisher:
Elsevier BV
Journal:
Procedia Computer Science
Conference/Event name:
International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2014
Issue Date:
13-Sep-2014
DOI:
10.1016/j.procs.2014.08.132
Type:
Conference Paper
ISSN:
18770509
Additional Links:
http://linkinghub.elsevier.com/retrieve/pii/S1877050914010977
Appears in Collections:
Conference Papers; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorHussain, Shahiden
dc.date.accessioned2015-05-06T13:23:09Zen
dc.date.available2015-05-06T13:23:09Zen
dc.date.issued2014-09-13en
dc.identifier.citationTotal Path Length and Number of Terminal Nodes for Decision Trees 2014, 35:514 Procedia Computer Scienceen
dc.identifier.issn18770509en
dc.identifier.doi10.1016/j.procs.2014.08.132en
dc.identifier.urihttp://hdl.handle.net/10754/552376en
dc.description.abstractThis paper presents a new tool for study of relationships between total path length (average depth) and number of terminal nodes for decision trees. These relationships are important from the point of view of optimization of decision trees. In this particular case of total path length and number of terminal nodes, the relationships between these two cost functions are closely related with space-time trade-off. In addition to algorithm to compute the relationships, the paper also presents results of experiments with datasets from UCI ML Repository1. These experiments show how two cost functions behave for a given decision table and the resulting plots show the Pareto frontier or Pareto set of optimal points. Furthermore, in some cases this Pareto frontier is a singleton showing the total optimality of decision trees for the given decision table.en
dc.publisherElsevier BVen
dc.relation.urlhttp://linkinghub.elsevier.com/retrieve/pii/S1877050914010977en
dc.rightsArchived with thanks to Procedia Computer Science. http://creativecommons.org/licenses/by-nc-nd/3.0/en
dc.subjectDecision treesen
dc.subjectnumber of terminal nodesen
dc.subjecttotal path lengthen
dc.subjectaverage depthen
dc.subjectPareto frontieren
dc.subjectPareto optimalen
dc.titleTotal Path Length and Number of Terminal Nodes for Decision Treesen
dc.typeConference Paperen
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.identifier.journalProcedia Computer Scienceen
dc.conference.date2014-09-15 to 2014-09-17en
dc.conference.nameInternational Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2014en
dc.conference.locationGdynia, POLen
dc.eprint.versionPublisher's Version/PDFen
kaust.authorHussain, Shahiden
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.