Revolutionizing brucellosis disease prediction with specialized machine learning techniques

Brucellosis, a pervasive infectious disease transmissible between animals and humans, poses a significant threat to public health and animal productivity worldwide. Traditional methods for predicting and controlling brucellosis have limitations in terms of accuracy and efficacy. This paper explores the utilization of machine learning (ML) techniques to revolutionize the approach to brucellosis prediction and management. Four distinct ML models, including Random Forest (RF), Radial Basis Function (RBF), Support Vector Machine (SVM), and J48 Trees, are evaluated based on various criteria such as correlation coefficient, root mean square error, mean absolute relative error, mean absolute error, Kappa statistic, and processing time. Among these models, RBF emerges as the most favorable choice, exhibiting the highest correlation coefficient and the lowest error metrics. It also maintains relatively efficient processing times. Additionally, the paper employs the Apriori algorithm for frequent item mining and association rule research within brucellosis datasets. The algorithm identifies robust associations, including relationships involving specific years, the non-pasteurized dairy product ratio, and brucellosis cases. Attribute ranking through ML reveals that average wind speed, non-pasteurized dairy product ratio, and rural ratio are crucial factors for brucellosis prediction. This study demonstrates the potential of machine learning in enhancing the accuracy and efficiency of brucellosis outbreak prediction. The Radial Basis Function model, in combination with insights from association rules and attribute ranking, can aid in the development of early warning systems and intervention strategies to mitigate the effect of Brucellosis on human health and animal populations.