<![CDATA[Forecasting sales time series data is essential for companies to support effective planning and decision-making processes. This study evaluates the strengths of the Seasonal Autoregressive Integrated Moving Average (SARIMA) and High-Order Fuzzy Time Series Chen (FTS Chen) models in predicting motorbike sales at Kalla Kars Company, a prominent automotive dealer in Sulawesi, Indonesia. SARIMA is renowned for accurately capturing seasonal patterns, while the FTS Chen model excels in handling data uncertainties and incorporating complex relationships through high-order fuzzy logic. Weekly sales data from January 2020 to February 2024 were analyzed, with 205 weeks used for training and 13 weeks for testing. The results indicate that the third-order FTS Chen model outperforms SARIMA, achieving a Root Mean Square Error (RMSE) of 1.88 and a Mean Absolute Percentage Error (MAPE) of 4.64%. Forecasts for the next eight weeks using the third-order FTS Chen model suggest a decline in sales, contrasting with the SARIMA model, which predicts a slight increase. These results show that Chen's FTS model is more accurate and reliable, making it an effective choice for forecasting Kalla Kars motorbike sales.]]>
Copyrights © 2024
