Optimization of Livestock Farming Using Multi-Tasking Regularisation Model

Effective management and monitoring of cow behavior attributes are vital for the success of dairy farming. This study introduces a comprehensive methodology that employs the Multi-Task Regularization (MTR) model to accurately predict three critical cow behavior attributes: activity level, feeding duration and rumination time. By integrating multiple machine learning algorithms within the MTR framework shared representations of input features are harnessed while task-specific output layers are maintained for each behavior attribute. Through rigorous experimentation the MTR model demonstrates superior performance surpassing other machine learning models in terms of accuracy as 89 per cent, 93 per cent and 91 per cent for the behavior attributes, precision and recall values across all behavior attributes. The precise predictions obtained from the MTR model empower farmers to make well-informed decisions pertaining to animal health, welfare and productivity enabling optimized feeding strategies, early identification of health concerns and enhanced herd management. Moreover, the potential integration of the MTR model with advanced monitoring technologies holds promise for comprehensive cow behavior prediction within the domain of dairy farming.