| dc.description.abstract | Excess inventory is a problem that is often faced in industry and manufacturing. High inventory conditions will result in high allocation of funds and will lead to increased costs in the company's operations. The condition of excess inventory in the last 2 years experienced by the feed mill company, the feed mill company is engaged in the animal feed production sector, which is located in Deli Serdang, North Sumatra, and markets its products in the regions of North Sumatra to Aceh and West Sumatra. Excess inventory that continues to occur for almost the last 2 years has the potential to cause losses in 1 year which reach 15 billion/year.
Based on the problem of potential losses caused, this study aims to design an inventory optimization model in which the optimization model will be collaborated with a machine learning approach. The design of the optimization model is prepared using a forecasting process using ARIMA (Autoregressive Integrated Moving Average). The machine learning data train used is a dateset from Actual Ordering, Forecast using ARIMA, re-order points, and carrying costs for 3 years, namely 2020, 2021 and 2022 which from the results of the evaluation of machine learning algorithm metrics that are suitable and effective for use in the dataset it is a support vector machine algorithm. The support vector machine algorithm can produce a more accurate prediction output by considering the pre-existing stock of animal feed packaging as a reference for determining animal feed packaging orders in the following month | en_US |
