Bayesian networks for mastitis management on dairy farms

Abstract

This manuscript presents the idea of providing dairy farmers with probability distributions to support decisions on mastitis management and illustrates its feasibility by two applications. Naive Bayesian networks were developed for both applications. The networks in the first application were used to compute probability distributions to support decisions on which cows from a mastitis alert list generated by detection sensors in an automatic milking system should be visually inspected for clinical mastitis. The computed probability distribution allows farmers to interpret the uncertainty in an alert. The network in the second application was used to compute probability distributions to support treatment decisions for clinical mastitis cases. The computed probability distribution allows for distinguishing a situation in which a single pathogen has a high probability from the situation where two or more pathogens have almost equal probabilities. The first situation would support the choice for a pathogen-specific treatment.