Analisis Implementasi K3 Untuk Meningkatkan Keselamatan Kerja di Bengkel PT. Kalitelu Teknik Menggunakan Metode HIRARC

Abstract

This study aims to analyze and compare the accuracy of the Double Moving Average (DMA) and Single Exponential Smoothing (SES) methods in forecasting the demand for fancy goldfish at the Wirajaya MSME. A descriptive-comparative quantitative approach was employed, utilizing historical monthly demand data for Ranchu, Oranda, and Lionhead goldfish from January to December 2025. The DMA method was tested using periods of n = 2, 3, and 4, while the SES method utilized smoothing parameters ranging from α = 0.1 to 0.9. Method accuracy was evaluated based on Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). The results indicate that DMA yielded MAPE values ​​of 5.97% for Ranchu, 4.91% for Oranda, and 6.47% for Lionhead. In contrast, SES produced MAPE values ​​of 10.33% for Ranchu (at α = 0.4), 7.58% for Oranda (at α = 0.5), and 10.14% for Lionhead (at α = 0.7). The comparison reveals that DMA achieved lower MAPE values ​​than SES across all three fish varieties. Consequently, DMA demonstrates superior forecasting accuracy and serves as a viable alternative for supporting production planning and inventory control of fancy goldfish at the Wirajaya MSME.