Machine Learning Based Prediction of Insufficient Herbage Allowance with Automated Feeding Behaviour and Activity Data

Abstract

Sensor technologies that measure grazing and ruminating behaviour as well as physical activities of individual cows are intended to be included in precision pasture management. One of the advantages of sensor data is they can be analysed to support farmers in many decision-making processes. This article thus considers the performance of a set of RumiWatchSystem recorded variables in the prediction of insufficient herbage allowance for spring calving dairy cows. Several commonly used models in machine learning (ML) were applied to the binary classification problem, i.e., sufficient or insufficient herbage allowance, and the predictive performance was compared based on the classification evaluation metrics. Most of the ML models and generalised linear model (GLM) performed similarly in leave-out-one-animal (LOOA) approach to validation studies. However, cross validation (CV) studies, where a portion of features in the test and training data resulted from the same cows, revealed that support vector machine (SVM), random forest (RF) and extreme gradient boosting (XGBoost) performed relatively better than other candidate models. In general, these ML models attained 88% AUC (area under receiver operating characteristic curve) and around 80% sensitivity, specificity, accuracy, precision and F-score. This study further identified that number of rumination chews per day and grazing bites per minute were the most important predictors and examined the marginal effects of the variables on model prediction towards a decision support system.

Review: Application and Prospective Discussion of Machine Learning for the Management of Dairy Farms

“…However, finding the right parameter or the right combination of parameters remains challenging [ 100 ]. Shafiullah et al [ 101 ] have taken their research a step further and were able to detect herbage shortages in the feeding and activity behaviors of grazing dairy cows and found that rumination chews per day and grazing bites per minute were the best predictors for insufficient grazing [ 101 ]. Machine learning models (support vector machine, random forest, and extreme gradient boosting) hereby performed better than the general linear model did in cross-validation [ 101 ].…”

Evaluation of a Binary Classification Approach to Detect Herbage Scarcity Based on Behavioral Responses of Grazing Dairy Cows

“…The study by Werner et al [ 17 ] was followed up by Shafiullah et al [ 19 ], who used their data to train models via supervised learning that evaluated sufficiency or insufficiency of herbage allowance on a 24 h basis in grazing dairy cows. The idea was to support farmers in scheduling paddock rotations before cow performance dropped.…”

“…The idea was to support farmers in scheduling paddock rotations before cow performance dropped. Using a binary system, Shafiullah et al [ 19 ] developed and validated several prediction approaches based on changes in the cows’ number of rumination chews per day and of their grazing bite frequency, along with six other variables. Among various statistical learning methods, a random forest model (RFM) performed best in classifying the binary feeding status of grazing dairy cows.…”

“…However, in order to correctly assess the generalizability of such models to different grazing conditions of dairy cows, and, thus, their possible implementation in future smart farming systems, they should be evaluated in commercial herds [ 6 , 20 ] and for a differing grazing system, such as a multi-day rotation. Variables selected and prediction models developed by Shafiullah et al [ 19 ] have not yet been evaluated using independent data from another dairy farm.…”

“…The aim of the present study was, therefore, to evaluate two prediction models recommended by Shafiullah et al [ 19 ] under commercial farming conditions. For this purpose, a practice-oriented rotational grazing system was set up with a gradually decreasing pasture allowance over six subsequent grazing days.…”