Discrimination of ingestive behavior in sheep using an electronic device based on a triaxial accelerometer and machine learning

• Published in: Computers and electronics in agriculture 2024-03, Vol.218, p.108657, Article 108657
• Main Authors: do Nascimento Amorim, Magno, Turco, Silvia Helena Nogueira, dos Santos Costa, Daniel, Ferreira, Iara Jeanice Souza, da Silva, Wedson Pereira, Sabino, Antonio Leopoldo Cardoso, da Silva-Miranda, Késia Oliveira
• Format: Article
• Language: English
• Subjects: Artificial intelligence; Classification models; Ovis aries; Ruminants; Sampling frequency
• ISSN: 0168-1699; 1872-7107
• DOI: 10.1016/j.compag.2024.108657

•We developed an electronic device to discriminate ingestive behavior in sheep.•We evaluated the device with different sampling frequencies and classification models.•ADXL345 accelerometer discriminates behavior feeding, rumination and idleness.•Random Forest was the best model for classifying ingestive behavior.•Sampling frequency of 0.05 Hz was accurate in classifying ingestive behavior. Avaluating ingestive behavior is crucial for making decisions about animal management; however, direct observation can be laborious, time-consuming, and exhaustive for the observer. In this context, the development of measurement devices coupled with computational methods to reliably discriminate animal behavior and identify an appropriate sampling frequency for battery autonomy without compromising precision is important. This study aimed to develop an electronic device for discriminating ingestive behavior in sheep using a triaxial accelerometer and machine learning, considering different sampling frequencies and classification models. The analyzed sampling frequencies were 1, 0.2, 0.1, 0.067, and 0.05 Hz, validated in 18 sheep. The electronic device successfully distinguished feeding, rumination, and other behaviors. The highest values in the results for ingestive behavior classification were achieved by the Random Forest model with and without variable selection, with accuracies ranging from 85.2 to 88.6 %, Kappa indices from 0.78 to 0.83, and false-positive rates from 5.4 to 7.2. Furthermore, the best-performing model differentiated feeding, rumination, and other behaviors with sensitivities ranging from 85.70 to 88.33 % and specificities from 92.71 to 94.11 %. Sampling frequencies of 1, 0.2, 0.1, and 0.05 Hz showed statistically similar means (p