Contagious Bovine Pleuropneumonia (CBPP) is a highly infectious disease that poses a significant threat to cattle populations, particularly in sub-Saharan Africa, where livestock is a critical economic and food security resource. Early detection and intervention are essential to mitigate its impact on cattle productivity and rural livelihoods. This study aims to develop a predictive model for CBPP using the Random Forest (RF) algorithm, a robust and scalable machine learning technique renowned for its accuracy and ability to handle complex datasets. The research utilized a dataset of 847 samples collected from veterinary clinics in Yola, Adamawa State, and online sources. Key symptoms of CBPP, such as respiratory distress, nasal discharge, fever, and pleuritis, served as input features for the model. Preprocessing steps, including outlier detection, handling missing values, and data normalization, were applied to ensure the dataset's quality and reliability. The RF algorithm was trained and validated to classify CBPP cases effectively, with performance metrics such as accuracy, precision, recall, and specificity used to evaluate its effectiveness. The experimental results demonstrated an impressive prediction accuracy of 95.65%. Results demonstrated the RF model's high accuracy and robustness in predicting CBPP, offering a practical tool for early diagnosis and decision-making. The study further highlights the potential of deploying the model in a mobile application, enabling cattle breeders to perform preliminary diagnoses and seek timely veterinary intervention. This approach aligns with global efforts to integrate technology into agricultural practices, enhancing livestock disease management and supporting sustainable agricultural development. The findings contribute to the growing body of knowledge on applying machine learning to livestock disease prediction, showcasing Random Forest's potential to transform disease management strategies in the livestock sector.
Read full abstract