Implementasi Simple Moving Average dan Rank Order Centroid–Weighted Moving Average untuk Peramalan Produksi Keripik Tempe UMKM X

Abstract

This study aims to forecast the production demand of tempeh chips at MSME X in Kubu Raya Regency, West Kalimantan. Forecasting was conducted using the Simple Moving Average (SMA) and Weighted Moving Average (WMA) methods with data periods (n) of 3 and 6 months. In the WMA method, weights were determined using the Rank Order Centroid (ROC) approach. The analyzed data consisted of tempeh chip production demand from December 2023 to November 2024, and the forecasting calculations were performed using Microsoft Excel. The accuracy of the forecasting results was evaluated using the Mean Absolute Deviation (MAD) metric. The results showed that the lowest MAD value was obtained using the WMA method with n = 6 (8.91), followed by SMA with n = 6 (10.94), SMA with n = 3 (17.78), and WMA with n = 3 (18.37). These findings indicate that the longer the historical data period used, the better the forecasting accuracy. Moreover, the application of ROC weighting in the WMA method enables predictions to be more responsive to actual trend changes. Based on these results, MSME X is recommended to adopt the WMA method with n = 6 for forecasting tempeh chip production demand.