<![CDATA[Broiler chicken farming faces challenges in predicting productivity due to production performance fluctuations and inefficient manual recording. This research aims to develop a broiler chicken productivity prediction system using the ARIMA method at Mitra Unggas Muara Bulian Chicken Farm. Research data was collected from daily production records for 30 days including population, mortality, average weight, and feed consumption. The method used is the Team Data Science Process (TDSP) to design a website-based prediction system. The research results show that the ARIMA (1,1,3) model can predict harvest weight for 1-3 days ahead with 86.96% accuracy, MAE 3,889.0 grams, and MAPE 13.04%. This system provides monitoring, prediction, and reporting features in real-time through easy-to-understand graphical visualizations. The implementation of this system is expected to support more effective farm management and assist farmers in making strategic decisions regarding production planning and optimal harvest timing ]]>
Copyrights © 2026
