<![CDATA[The optimal availability of pharmaceutical raw materials is a vital aspect in ensuring the continuity of production within the pharmaceutical industry. PT. Seikyo Indochem faces challenges in accurately forecasting raw material requirements due to the fluctuating and complex nature of the data. This study implements the Wavelet Transform method combined with ARIMA (Auto-Regressive Integrated Moving Average) to enhance the accuracy of demand forecasting. Wavelet Transform is utilized to decompose historical data into low- and high-frequency components, enabling a more in-depth analysis of seasonal patterns and trends. The low-frequency component is analyzed using ARIMA to predict long-term patterns, while the high-frequency component is used to capture short-term fluctuations. The results show that this hybrid approach reduces the prediction error (Mean Absolute Percentage Error) by 15 percent compared to using ARIMA alone. This model provides a more reliable predictive solution to support efficient inventory management of pharmaceutical raw materials.]]>
Copyrights © 2025
