An empirical comparison among quality measures for pattern based classifiers

Octavio Loyola-González, Milton Garciá-Borroto, José Fco Martínez-Trinidad, Jesús Ariel Carrasco-Ochoa

Research output: Contribution to conferencePaper

5 Citations (Scopus)

Abstract

© 2014 IOS Press and the authors. All rights reserved. Measuring the quality of a contrast pattern is an active and relevant area of pattern recognition and data mining. Quality measures are important tools in very different scenarios like supervised classification, pattern based clustering, and association rule mining. Consequently, and due to the large collection of available measures, it is important to perform comparative studies for each particular context. Most published studies comparing quality measures are theoretical and in the context of association rule evaluation. In this paper, we present an empirical comparison of the behavior of 33 quality measures in the context of supervised classification and contrast pattern filtering. A comprehensive experimentation using several databases compares the behavior of these measures in three different contexts: as aggregation value, as pattern evaluation for classification, and as pattern evaluation for filtering. Experiments also show that top-accurate quality measures for classification have a deceptive performance for pattern filtering, because they cannot distinguish among patterns with zero support in the negative class.
Original languageAmerican English
PagesS5-S17
Number of pages0
DOIs
StatePublished - 1 Jan 2014
Externally publishedYes
EventIntelligent Data Analysis -
Duration: 1 Jan 2017 → …

Conference

ConferenceIntelligent Data Analysis
Period1/01/17 → …

Fingerprint

Classifiers
Association rules
Pattern recognition
Data mining
Agglomeration
Experiments

Cite this

Loyola-González, O., Garciá-Borroto, M., Martínez-Trinidad, J. F., & Carrasco-Ochoa, J. A. (2014). An empirical comparison among quality measures for pattern based classifiers. S5-S17. Paper presented at Intelligent Data Analysis, . https://doi.org/10.3233/IDA-140705
Loyola-González, Octavio ; Garciá-Borroto, Milton ; Martínez-Trinidad, José Fco ; Carrasco-Ochoa, Jesús Ariel. / An empirical comparison among quality measures for pattern based classifiers. Paper presented at Intelligent Data Analysis, .
@conference{1b31ae8aedc4436ab723e21516cb7295,
title = "An empirical comparison among quality measures for pattern based classifiers",
abstract = "{\circledC} 2014 IOS Press and the authors. All rights reserved. Measuring the quality of a contrast pattern is an active and relevant area of pattern recognition and data mining. Quality measures are important tools in very different scenarios like supervised classification, pattern based clustering, and association rule mining. Consequently, and due to the large collection of available measures, it is important to perform comparative studies for each particular context. Most published studies comparing quality measures are theoretical and in the context of association rule evaluation. In this paper, we present an empirical comparison of the behavior of 33 quality measures in the context of supervised classification and contrast pattern filtering. A comprehensive experimentation using several databases compares the behavior of these measures in three different contexts: as aggregation value, as pattern evaluation for classification, and as pattern evaluation for filtering. Experiments also show that top-accurate quality measures for classification have a deceptive performance for pattern filtering, because they cannot distinguish among patterns with zero support in the negative class.",
author = "Octavio Loyola-Gonz{\'a}lez and Milton Garci{\'a}-Borroto and Mart{\'i}nez-Trinidad, {Jos{\'e} Fco} and Carrasco-Ochoa, {Jes{\'u}s Ariel}",
year = "2014",
month = "1",
day = "1",
doi = "10.3233/IDA-140705",
language = "American English",
pages = "S5--S17",
note = "Intelligent Data Analysis ; Conference date: 01-01-2017",

}

Loyola-González, O, Garciá-Borroto, M, Martínez-Trinidad, JF & Carrasco-Ochoa, JA 2014, 'An empirical comparison among quality measures for pattern based classifiers', Paper presented at Intelligent Data Analysis, 1/01/17 pp. S5-S17. https://doi.org/10.3233/IDA-140705

An empirical comparison among quality measures for pattern based classifiers. / Loyola-González, Octavio; Garciá-Borroto, Milton; Martínez-Trinidad, José Fco; Carrasco-Ochoa, Jesús Ariel.

2014. S5-S17 Paper presented at Intelligent Data Analysis, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - An empirical comparison among quality measures for pattern based classifiers

AU - Loyola-González, Octavio

AU - Garciá-Borroto, Milton

AU - Martínez-Trinidad, José Fco

AU - Carrasco-Ochoa, Jesús Ariel

PY - 2014/1/1

Y1 - 2014/1/1

N2 - © 2014 IOS Press and the authors. All rights reserved. Measuring the quality of a contrast pattern is an active and relevant area of pattern recognition and data mining. Quality measures are important tools in very different scenarios like supervised classification, pattern based clustering, and association rule mining. Consequently, and due to the large collection of available measures, it is important to perform comparative studies for each particular context. Most published studies comparing quality measures are theoretical and in the context of association rule evaluation. In this paper, we present an empirical comparison of the behavior of 33 quality measures in the context of supervised classification and contrast pattern filtering. A comprehensive experimentation using several databases compares the behavior of these measures in three different contexts: as aggregation value, as pattern evaluation for classification, and as pattern evaluation for filtering. Experiments also show that top-accurate quality measures for classification have a deceptive performance for pattern filtering, because they cannot distinguish among patterns with zero support in the negative class.

AB - © 2014 IOS Press and the authors. All rights reserved. Measuring the quality of a contrast pattern is an active and relevant area of pattern recognition and data mining. Quality measures are important tools in very different scenarios like supervised classification, pattern based clustering, and association rule mining. Consequently, and due to the large collection of available measures, it is important to perform comparative studies for each particular context. Most published studies comparing quality measures are theoretical and in the context of association rule evaluation. In this paper, we present an empirical comparison of the behavior of 33 quality measures in the context of supervised classification and contrast pattern filtering. A comprehensive experimentation using several databases compares the behavior of these measures in three different contexts: as aggregation value, as pattern evaluation for classification, and as pattern evaluation for filtering. Experiments also show that top-accurate quality measures for classification have a deceptive performance for pattern filtering, because they cannot distinguish among patterns with zero support in the negative class.

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84920754906&origin=inward

UR - https://www.scopus.com/inward/citedby.uri?partnerID=HzOxMe3b&scp=84920754906&origin=inward

U2 - 10.3233/IDA-140705

DO - 10.3233/IDA-140705

M3 - Paper

SP - S5-S17

ER -

Loyola-González O, Garciá-Borroto M, Martínez-Trinidad JF, Carrasco-Ochoa JA. An empirical comparison among quality measures for pattern based classifiers. 2014. Paper presented at Intelligent Data Analysis, . https://doi.org/10.3233/IDA-140705