Forecasting Beef Production with Comparison of Linear Regression and DMA Methods Based on n-th Ordo 3

Abstract

Beef is considered a high-value commodity because it is an important food source of protein. Interest in beef is increasing along with increasing people's incomes and awareness of the importance of fulfilling nutrition. Demand for beef is expected to continue to increase. According to the Central Statistics Agency (CSA), beef production in Jakarta shows an increasing trend every year. In the last 10 years, beef production has increased significantly, but in 2020 there was a decrease in production of 7,240.68 tons due to the lockdown due to the corona virus outbreak. After that, in 2021, production reached 16,381.81 tons and will continue to increase in 2022 and 2023. Based on the above phenomenon, the aim of this research is to support the success and sustainability of the beef industry by ensuring that supply matches demand, resources are used optimally, and risks can be managed well. To predict beef production, an accurate method, model or approach is needed. One way to predict beef production in Jakarta is to use the Linear Regression and Double Moving Average (DMA) methodsThe way the Linear Regression and DMA methods work is to forecast based on concepts and properties. The concepts and properties of Linear Regression are models, functions, estimates and forecasting results, while DMA performs time series analysis based on moving averages. After analysis using MAPE, it was found that the algorithm that had the smallest error value was the linear regression algorithm with a percentage for the monthly period of 15% while for the year period it was 17% compared to DMA. So in this case it would be very appropriate to use the Linear Regression method from the error values obtained.

Keywords

• forecasting
• beef
• production
• linear regression
• double moving average.