The Early Prediction of Common Disorders in Dairy Cows Monitored by Automatic Systems with Machine Learning Algorithms

SIMPLE SUMMARY: Identifying cows with a higher risk of health disorders such as clinical mastitis, subclinical ketosis, lameness, and metritis could be advantageous for farms to prevent and ameliorate the negative effects of these disorders in a timely manner. In this study, we adopt eight machine learning algorithms using an R software for analyzing a dataset of 14-dimensions of dairy cows with health disorders across the whole lactation period in intensive Chinese dairy farms, applying automatic monitoring systems and milking systems. The variables analyzed by the machine learning algorithms include milk yield, physical activity, changes in rumination time, and the electrical conductivity of milk. Six parameters were presented to evaluate the performance metrics of the models, with the Rpart algorithm outperforming others and indicating a strong generalization ability of this algorithm. A total of 10 variables of greater importance in three models of Rpart, eXtreme Gradient, and Adaboost demonstrated the consistency of those variables as predictors for disorders of dairy cows monitored by automatic systems. The results obtained in this study highlighted the importance of using big data on the farm to develop predictive and prescriptive decision support tools to boost the development of precision livestock farming. ABSTRACT: We use multidimensional data from automated monitoring systems and milking systems to predict disorders of dairy cows by employing eight machine learning algorithms. The data included the season, days in milking, parity, age at the time of disorders, milk yield (kg/day), activity (unitless), six variables related to rumination time, and two variables related to the electrical conductivity of milk. We analyze 131 sick cows and 149 healthy cows with identical lactation days and parity; all data are collected on the same day, which corresponds to the diagnosis day for disordered cows. For disordered cows, each variable, except the ratio of rumination time from daytime to nighttime, displays a decreasing/increasing trend from d-7 or d-3 to d0 and/or d-1, with the d0, d-1, or d-2 values reaching the minimum or maximum. The test data sensitivity for three algorithms exceeded 80%, and the accuracies of the eight algorithms ranged from 65.08% to 84.21%. The area under the curve (AUC) of the three algorithms was >80%. Overall, Rpart best predicts the disorders with an accuracy, precision, and AUC of 81.58%, 92.86%, and 0.908, respectively. The machine learning algorithms may be an appropriate and powerful decision support and monitoring tool to detect herds with common health disorders.