<![CDATA[The distribution of cattle before Eid al-Adha often leads to transport-induced stress, negatively affecting livestock performance and economic value. This study aims to develop a predictive model of post-transport cattle performance using Artificial Neural Networks (ANN). The dataset includes physiological parameters (rectal temperature, heart rate, and respiration) and blood metabolites (glucose and creatinine) collected before and after transportation. Data augmentation and feature selection were applied using Pearson correlation to address class imbalance. The ANN model was tuned with regularisation and dropout techniques to prevent overfitting. Evaluation results show that the model achieved 91% accuracy, with F1-scores of 0.90 (Increase), 0.97 (Stable), and 0.87 (Decrease). These findings demonstrate that ANN can capture complex patterns of physiological conditions in cattle and provide reliable predictions. This model has the potential to serve as the basis for developing an early warning system to minimize the risk of performance decline in cattle due to transport stress more adaptively and efficiently.]]>
Copyrights © 2025
