Comparing supervised learning methods for classifying sex, age, context and individual Mudi dogs from barking

Ana Larrañaga, Concha Bielza, P. Pongrácz, Tamás Faragó, Anna Bálint, Pedro Larrañaga

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Barking is perhaps the most characteristic form of vocalization in dogs; however, very little is known about its role in the intraspecific communication of this species. Besides the obvious need for ethological research, both in the field and in the laboratory, the possible information content of barks can also be explored by computerized acoustic analyses. This study compares four different supervised learning methods (naive Bayes, classification trees, k-nearest neighbors and logistic regression) combined with three strategies for selecting variables (all variables, filter and wrapper feature subset selections) to classify Mudi dogs by sex, age, context and individual from their barks. The classification accuracy of the models obtained was estimated by means of k-fold cross-validation. Percentages of correct classifications were 85.13 % for determining sex, 80.25 % for predicting age (recodified as young, adult and old), 55.50 % for classifying contexts (seven situations) and 67.63 % for recognizing individuals (8 dogs), so the results are encouraging. The best-performing method was k-nearest neighbors following a wrapper feature selection approach. The results for classifying contexts and recognizing individual dogs were better with this method than they were for other approaches reported in the specialized literature. This is the first time that the sex and age of domestic dogs have been predicted with the help of sound analysis. This study shows that dog barks carry ample information regarding the caller’s indexical features. Our computerized analysis provides indirect proof that barks may serve as an important source of information for dogs as well.

Original languageEnglish
Pages (from-to)405-421
Number of pages17
JournalAnimal Cognition
Volume18
Issue number2
DOIs
Publication statusPublished - 2014

Fingerprint

learning
Learning
Dogs
dogs
gender
bark
methodology
vocalization
information sources
supervised learning
dog
method
young adults
Acoustics
animal communication
acoustics
Young Adult
logistics
Logistic Models
Communication

Keywords

  • Acoustic communication
  • Feature subset selection
  • K-fold cross-validation
  • Machine learning
  • Mudi dog barks
  • Supervised classification

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics
  • Experimental and Cognitive Psychology

Cite this

Comparing supervised learning methods for classifying sex, age, context and individual Mudi dogs from barking. / Larrañaga, Ana; Bielza, Concha; Pongrácz, P.; Faragó, Tamás; Bálint, Anna; Larrañaga, Pedro.

In: Animal Cognition, Vol. 18, No. 2, 2014, p. 405-421.

Research output: Contribution to journalArticle

Larrañaga, Ana ; Bielza, Concha ; Pongrácz, P. ; Faragó, Tamás ; Bálint, Anna ; Larrañaga, Pedro. / Comparing supervised learning methods for classifying sex, age, context and individual Mudi dogs from barking. In: Animal Cognition. 2014 ; Vol. 18, No. 2. pp. 405-421.
@article{568b775a4e544558bbeaa6afb21d7575,
title = "Comparing supervised learning methods for classifying sex, age, context and individual Mudi dogs from barking",
abstract = "Barking is perhaps the most characteristic form of vocalization in dogs; however, very little is known about its role in the intraspecific communication of this species. Besides the obvious need for ethological research, both in the field and in the laboratory, the possible information content of barks can also be explored by computerized acoustic analyses. This study compares four different supervised learning methods (naive Bayes, classification trees, k-nearest neighbors and logistic regression) combined with three strategies for selecting variables (all variables, filter and wrapper feature subset selections) to classify Mudi dogs by sex, age, context and individual from their barks. The classification accuracy of the models obtained was estimated by means of k-fold cross-validation. Percentages of correct classifications were 85.13 {\%} for determining sex, 80.25 {\%} for predicting age (recodified as young, adult and old), 55.50 {\%} for classifying contexts (seven situations) and 67.63 {\%} for recognizing individuals (8 dogs), so the results are encouraging. The best-performing method was k-nearest neighbors following a wrapper feature selection approach. The results for classifying contexts and recognizing individual dogs were better with this method than they were for other approaches reported in the specialized literature. This is the first time that the sex and age of domestic dogs have been predicted with the help of sound analysis. This study shows that dog barks carry ample information regarding the caller’s indexical features. Our computerized analysis provides indirect proof that barks may serve as an important source of information for dogs as well.",
keywords = "Acoustic communication, Feature subset selection, K-fold cross-validation, Machine learning, Mudi dog barks, Supervised classification",
author = "Ana Larra{\~n}aga and Concha Bielza and P. Pongr{\'a}cz and Tam{\'a}s Farag{\'o} and Anna B{\'a}lint and Pedro Larra{\~n}aga",
year = "2014",
doi = "10.1007/s10071-014-0811-7",
language = "English",
volume = "18",
pages = "405--421",
journal = "Animal Cognition",
issn = "1435-9448",
publisher = "Springer Verlag",
number = "2",

}

TY - JOUR

T1 - Comparing supervised learning methods for classifying sex, age, context and individual Mudi dogs from barking

AU - Larrañaga, Ana

AU - Bielza, Concha

AU - Pongrácz, P.

AU - Faragó, Tamás

AU - Bálint, Anna

AU - Larrañaga, Pedro

PY - 2014

Y1 - 2014

N2 - Barking is perhaps the most characteristic form of vocalization in dogs; however, very little is known about its role in the intraspecific communication of this species. Besides the obvious need for ethological research, both in the field and in the laboratory, the possible information content of barks can also be explored by computerized acoustic analyses. This study compares four different supervised learning methods (naive Bayes, classification trees, k-nearest neighbors and logistic regression) combined with three strategies for selecting variables (all variables, filter and wrapper feature subset selections) to classify Mudi dogs by sex, age, context and individual from their barks. The classification accuracy of the models obtained was estimated by means of k-fold cross-validation. Percentages of correct classifications were 85.13 % for determining sex, 80.25 % for predicting age (recodified as young, adult and old), 55.50 % for classifying contexts (seven situations) and 67.63 % for recognizing individuals (8 dogs), so the results are encouraging. The best-performing method was k-nearest neighbors following a wrapper feature selection approach. The results for classifying contexts and recognizing individual dogs were better with this method than they were for other approaches reported in the specialized literature. This is the first time that the sex and age of domestic dogs have been predicted with the help of sound analysis. This study shows that dog barks carry ample information regarding the caller’s indexical features. Our computerized analysis provides indirect proof that barks may serve as an important source of information for dogs as well.

AB - Barking is perhaps the most characteristic form of vocalization in dogs; however, very little is known about its role in the intraspecific communication of this species. Besides the obvious need for ethological research, both in the field and in the laboratory, the possible information content of barks can also be explored by computerized acoustic analyses. This study compares four different supervised learning methods (naive Bayes, classification trees, k-nearest neighbors and logistic regression) combined with three strategies for selecting variables (all variables, filter and wrapper feature subset selections) to classify Mudi dogs by sex, age, context and individual from their barks. The classification accuracy of the models obtained was estimated by means of k-fold cross-validation. Percentages of correct classifications were 85.13 % for determining sex, 80.25 % for predicting age (recodified as young, adult and old), 55.50 % for classifying contexts (seven situations) and 67.63 % for recognizing individuals (8 dogs), so the results are encouraging. The best-performing method was k-nearest neighbors following a wrapper feature selection approach. The results for classifying contexts and recognizing individual dogs were better with this method than they were for other approaches reported in the specialized literature. This is the first time that the sex and age of domestic dogs have been predicted with the help of sound analysis. This study shows that dog barks carry ample information regarding the caller’s indexical features. Our computerized analysis provides indirect proof that barks may serve as an important source of information for dogs as well.

KW - Acoustic communication

KW - Feature subset selection

KW - K-fold cross-validation

KW - Machine learning

KW - Mudi dog barks

KW - Supervised classification

UR - http://www.scopus.com/inward/record.url?scp=84922666162&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84922666162&partnerID=8YFLogxK

U2 - 10.1007/s10071-014-0811-7

DO - 10.1007/s10071-014-0811-7

M3 - Article

C2 - 25308549

AN - SCOPUS:84922666162

VL - 18

SP - 405

EP - 421

JO - Animal Cognition

JF - Animal Cognition

SN - 1435-9448

IS - 2

ER -