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 A. Larranaga Student at the Universidad Alfonso X El Sabio, Av. Universidad, 1, 28691 Villanueva de la Canada, Madrid, Spain 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.

The bark, the howl and the bark-howl: Identity cues in dingoes’ multicomponent calls

“…Contrary to our second prediction, bark segments were also individually distinctive, although to a lesser extent than howl segments. The 59% classification success is lower than that reported for coyote barks (70%), but similar to rates achieved in domestic dogs, which range from 40% to 68% depending on the methodologies and sample sizes (Yin and McCowan, 2004;Mitchell et al, 2006;Molnár et al, 2008;Larranaga et al, 2014).…”

Automatic Classification of Context in Induced Barking

“…When we evaluated by Resample, we obtained an F-measure of 0.69 (as shown in Table 2) and an accuracy of 68.64%. This represents an improvement on the results reported by Larrañaga et al [6] using the same data and the same evaluation scheme. They obtained an accuracy of 55.50% using a k-nearest neighbour classifier and a wrapper feature selection method.…”

“…Classification accuracy is evaluated using this reduced dataset and 10FCV. We included this validation scheme in order to compare our results with the reported by [6] where they used the same data and similar method for re-sampling. Leave One Dog Out Validation (LODOV) with the objective of measure the impact and dependency of individuals in the classification.…”

“…They constructed a set of acoustic descriptors using an evolutionary algorithm and feature selection techniques. Larrañaga et al [6] compared several supervised machine learning methods for four classification tasks: sex, age, context and individual. They tested four machine learning methods and a set of 29 acoustic measures extracted from each barking recording.…”

A bark of its own kind – the acoustics of ‘annoying’ dog barks suggests a specific attention-evoking effect for humans

“…In previous studies it was found that specific acoustic features of dog barking (e.g. fundamental frequency, tonality and interbark-intervals) play an important role in interspecific communication, as they convey contextual and emotional information and even individual characteristicssuch as age or sexof the dog (Yin 2002;Yin and McCowan 2004;Molnár et al 2010;Pongrácz et al 2010Pongrácz et al , 2011Larrañaga et al 2015). Compared to their wild relatives (e.g.…”