Indonesia's Breakthrough in Optimized Yarn Forecasting for Textile Demand Accuracy

Abstract

PT. XY, a textile company specializing in woven sarongs, faces fluctuating demand during Islamic religious celebrations, impacting production. In Ramadhan 2023, production increased by 30%, but warp yarn availability was insufficient. This study forecasts warp yarn production over twelve periods, comparing Double Exponential Smoothing Holt’s (DES) and Holt-Winter’s Exponential Smoothing (WES) methods, optimized using the golden section method. Using historical data from January 2021 to April 2023, WES with golden section parameters (α1 = 0.67387, β1 = 0.08756, γ2 = 0.85408) achieved the best accuracy with a MAPE of 5.5437%. The WES method is recommended for improving production planning at PT. XY, with future research suggested to explore production correlations and procurement costs. Highlight: Demand Fluctuation: PT. XY experiences significant demand changes during Islamic religious celebrations. Forecasting Methods: Comparing DES and WES methods for predicting warp yarn production. Optimal Accuracy: WES with golden section optimization achieved the lowest MAPE of 5.5437%.   Keywoard: Textile Industry, Warp yarn forecasting, Production Planning, Holt-Winter's method, Golden section optimization