The Use of Triaxial Accelerometers and Machine Learning Algorithms for Behavioural Identiﬁcation in Domestic Dogs (Canis familiaris) - A Validation Study

Q: What problem does the study address about dog behavioural assessment?

Conventional assessment methods are time consuming, labour intensive, and prone to bias, which makes large-scale and rapid implementation difficult.

Q: What device and method are validated in the study?

The study validates the ActiGraph® accelerometer as a tool for behavioural classification using a machine learning approach.

Q: How accurate is the machine learning model for identifying dog behaviours?

The model identified nine dog behaviours with an overall accuracy of 74%, with per-behaviour ranges from 54 to 93%.

Aug 5, 2026

Upload by Levi

Assessing the behaviour and physical attributes of domesticated dogs supports decisions for companionship and specific roles such as hunting, military, or service. Traditional behavioural assessment approaches can be time consuming, labour intensive, and vulnerable to bias, limiting large-scale and fast deployment. This validation study evaluates the ActiGraph® tri-axial accelerometer for behavioural classification using machine learning to identify nine behaviours with an overall accuracy of 74%, and shows dynamic body acceleration correlates with time spent on active behaviours.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages20

Reading Time50 min

AccessFree to download

Introduction

Behavioural assessment needs

Role of accelerometers

Machine learning for behaviour modeling

sensors Article The Use of Triaxial Accelerometers and Machine Learning Algorithms for Behavioural Identiﬁcation in Domestic Dogs (Canis familiaris): A Validation Study Cushla Redmond, Michelle Smit , Ina Draganova , Rene Corner-Thomas , David Thomas * and Christopher Andrews Citation: Redmond, C.; Smit, M.; Draganova, I.; Corner-Thomas, R.; Thomas, D.; Andrews, C. The Use of Triaxial Accelerometers and Machine Learning Algorithms for Behavioural Identiﬁcation in Domestic Dogs (Canis familiaris): A Validation Study. Sensors 2024, 24, 5955. [https://](https://) [doi.org/10.3390/s24185955](doi.org/10.3390/s24185955)[ ](doi.org/10.3390/s24185955)Academic Editors: David T. Fullwood and Anton E. Bowden Received: 26 June 2024 Revised: 4 September 2024 Accepted: 10 September 2024 Published: 13 September 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://](https://)[ ](https://)[creativecommons.org/licenses/by/](creativecommons.org/licenses/by/)[ ](creativecommons.org/licenses/by/)[4.0/](4.0/)) . School of Agriculture and Environment, Massey University, Palmerston North 4410, New Zealand; [cushla-r-@hotmail.com](cushla-r-@hotmail.com) (C.R.); [m.smit@massey.ac.nz](m.smit@massey.ac.nz) (M.S.); [i.draganova@massey.ac.nz](i.draganova@massey.ac.nz) (I.D.); [r.corner@massey.ac.nz](r.corner@massey.ac.nz) (R.C.-T.); [c.j.andrews@massey.ac.nz](c.j.andrews@massey.ac.nz) (C.A.) * Correspondence: [d.g.thomas@massey.ac.nz](d.g.thomas@massey.ac.nz) Abstract: Assessing the behaviour and physical attributes of domesticated dogs is critical for predicting the suitability of animals for companionship or speciﬁc roles such as hunting, military or service. Common methods of behavioural assessment can be time consuming, labour-intensive, and subject tobias, making large-scale and rapid implementation challenging. Objective, practical and time effective behaviour measures may be facilitated by remote and automated devices such as accelerometers. This study, therefore, aimed to validate the ActiGraph® accelerometer as a tool for behavioural classiﬁcation. This study used a machine learning method that identiﬁed nine dog behaviours with an overall accuracy of 74%(range for each behaviour was 54 to 93%) . In addition, overall body dynamic acceleration was found to be correlated with the amount of time spent exhibiting active behaviours (barking, locomotion, scratching, snifﬁng, and standing; R2 = 0 .91, p < 0 .001) . Machine learning was an effective method to build a model to classify behaviours such as barking, defecating, drinking, eating, locomotion, resting-asleep, resting-alert, snifﬁng, and standing with high overall accuracy whilst maintaining a large behavioural repertoire. Keywords: algorithm; behaviour classiﬁcation; overall activity; random forest 1. Introduction Domestic dogs (Canis familiaris) have become an integral part of human society, serving various purposes such as companionship, working, hunting, research, and military service roles [1,2] . Assessing their behaviour is critical for predicting suitability and targeting speciﬁc temperament characteristics required for these speciﬁc roles [3] . Importantly, behavioural assessments not only facilitate selection of desirable traits, but also enhance our understanding of the overall welfare state of dogs [4–6] . Common methods of behavioural assessment include test batteries, observational methods, and questionnaires [7–9] . These methods, however, can be time consuming, labour intensive, and subject to bias, making large-scale implementation challenging [10–15] . There is a need for methods of behavioural assessment that are objective, practical and do not require extensive time or training from human observers [16] . Recent technological advances have facilitated remote and automated b

FAQs

What problem does the study address about dog behavioural assessment?

Conventional assessment methods are time consuming, labour intensive, and prone to bias, which makes large-scale and rapid implementation difficult.

What device and method are validated in the study?

The study validates the ActiGraph® accelerometer as a tool for behavioural classification using a machine learning approach.

How accurate is the machine learning model for identifying dog behaviours?

The model identified nine dog behaviours with an overall accuracy of 74%, with per-behaviour ranges from 54 to 93%.

Aug 5, 2026

Upload by Levi