Classification of broiler behaviours using triaxial accelerometer and machine learning.

Area of Science:

• Animal Science
• Machine Learning
• Biotechnology

Background:

• Understanding broiler chicken behavior is crucial for animal welfare and effective farm management.
• Wearable sensors offer a non-invasive method for monitoring animal activity.

Purpose of the Study:

• To classify specific broiler behaviors using data from wearable accelerometers.
• To compare the performance of K-Nearest Neighbour (KNN) and Support Vector Machine (SVM) machine learning models for behavior classification.

Main Methods:

• Triaxial accelerometers recorded broiler accelerations at 40 Hz.
• Data were labeled for four behaviors: walking, resting, feeding, and drinking.
• Extracted motion features were analyzed using machine learning models with varying window lengths.

Main Results:

• Both KNN and SVM models achieved high sensitivity for resting and walking behaviors.
• SVM demonstrated higher accuracy than KNN in differentiating feeding and drinking behaviors.
• A 1-second sliding window and increased training data improved classification performance.

Conclusions:

• Machine learning analysis of accelerometer data can effectively classify broiler behaviors.
• Different machine learning models exhibit varying performance in classifying specific broiler behaviors.
• This technology has implications for improving broiler welfare and farm management strategies.