The effect of an accumulation algorithm on the predictive accuracy of ARIMA models

This study explores the use of the autoregressive integrated moving average (ARIMA) data-driven modeling approach for forecasting peanut yields in Sudan. Two tests were conducted: one using the original dataset and another using accumulated data. The main objective was to improve forecasting accuracy by applying a method that incorporates accumulated data for future predictions. The results, based on a comparison of the two tests, indicate that the proposed approach enhances prediction clarity. Model identification showed an increase in the coefficient of determination, a decrease in the Bayesian information criterion (BIC), and a reduction in the mean absolute error. These outcomes suggest that the proposed method may provide more accurate forecasts and could be useful for forecasting in various fields.