Goats on the Move: Evaluating Machine Learning Models for Goat Activity Analysis Using Accelerometer Data

Abstract

Simple Summary This study developed and investigated a method to monitor animal activity by placing sensors on goats to better understand their behavior and improve their living conditions. Previous methods struggled to accurately classify complex movement data, but recent advancements in deep learning have made this task more effective. We compared the performance of three machine learning models: a basic neural network, a Convolutional Neural Network, and a hybrid model combining both. We evaluated two preprocessing methods to make the sensor data invariant to rotations, which could occur due to different placements of the sensors or slipping of the collars. Our results showed that the hybrid model, especially with these rotation invariant transformations, was the most accurate in identifying different goat activities. This research can help farmers monitor the health and well-being of their goats more effectively, ensuring better living conditions and improving dairy production. Our findings highlight the potential of advanced data analysis to enhance animal welfare and farm management practices.Abstract Putting sensors on the bodies of animals to automate animal activity recognition and gain insight into their behaviors can help improve their living conditions. Although previous hard-coded algorithms failed to classify complex time series obtained from accelerometer data, recent advances in deep learning have improved the task of animal activity recognition for the better. However, a comparative analysis of the generalizing capabilities of various models in combination with different input types has yet to be addressed. This study experimented with two techniques for transforming the segmented accelerometer data to make them more orientation-independent. The methods included calculating the magnitude of the three-axis accelerometer vector and calculating the Discrete Fourier Transform for both sets of three-axis data as the vector magnitude. Three different deep learning models were trained on this data: a Multilayer Perceptron, a Convolutional Neural Network, and an ensemble merging both called a hybrid Convolutional Neural Network. Besides mixed cross-validation, every model and input type combination was assessed on a goat-wise leave-one-out cross-validation set to evaluate its generalizing capability. Using orientation-independent data transformations gave promising results. A hybrid Convolutional Neural Network with L2-norm as the input combined the higher classification accuracy of a Convolutional Neural Network with the lower standard deviation of a Multilayer Perceptron. Most of the misclassifications occurred for behaviors that display similar accelerometer traces and minority classes, which could be improved in future work by assembling larger and more balanced datasets.