Replication Data for: Comparative performance analysis of three machine learning algorithms applied to sensor data in dairy cattle to predict metritis events II. Behaviors measured with a leg-attached accelerometer

Dataset containing all the performance metrics (sensitivity, specificity, positive and negative predictive values, accuracy, and F1 score) using a rank-based approach and three different decision thresholds (highest 20%, 30%, and 40% class probabilities). Each row is a model that is defined by the combination of classifier (k-NN, RF, SVM), type of behavior measured with the sensor device (lying time, lying bouts, steps, intake, intake visits), time of the day corresponding to the sensor data used (all day, evening-night), parity (all cows regardless of parity, primiparous, multiparous), time window to aggregate sensor data (24, 12, 6, 3 hours), and number of time steps before a given event used as model inputs as a function of the time window used to aggregate sensor data. Also, information regarding number of metritis events, number of non-metritis events, and total number of events for each model was recorded.