COMPARISON OF PSO AND ABC IN CHENG FUZZY TIME SERIES FOR RICE PRICE FORECASTING

Abstract

Rice prices as a primary food commodity in Indonesia play an important role in maintaining economic stability and public welfare, but tend to fluctuate, thus requiring accurate forecasting methods to support decision-making. Research on optimization in the Cheng Fuzzy Time Series (FTS Cheng) method remains limited, particularly in comparing the performance of Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) in rice price forecasting. This study aims to compare the performance of PSO and ABC optimization in the FTS Cheng method using monthly data from January 2018 to October 2025, with accuracy evaluated using MAE, RMSE, and MAPE. The forecasting process is carried out through interval formation on training and testing data to obtain an optimal model. The results show that FTS Cheng-ABC performs better, with an MAE of 97.947, RMSE of 142.855, and MAPE of 0.633%, compared to FTS Cheng-PSO with an MAE of 118.579, RMSE of 153.354, and MAPE of 0.767%. However, this study is limited to the use of the Fuzzy Time Series Cheng method with two optimization algorithms, namely PSO and ABC, and does not incorporate adaptive parameter mechanisms or comparisons with more advanced methods. Therefore, the FTS Cheng-ABC method is more effective and can be used to support policy decision-making related to rice price stability. This study contributes by providing a comparative analysis of PSO and ABC optimization in improving the performance of the FTS Cheng method for rice price forecasting in Indonesia.