Abstract
Inventory management is a major challenge in the clothing retail business. Toko Saing as a clothing retail business still performs manual stock recording, causing risks of overstock and stockout. This study aims to apply the Random Forest algorithm to predict stock availability and optimize marketing of clothing products. Historical sales data for 6 months (October 2024 - March 2025) was used as the training dataset. The prediction system was built web-based using PHP and MySQL. The test results show that the Random Forest algorithm is able to predict stock needs with an average accuracy of 77.3% (R²) and Mean Absolute Error (MAE) of 0.6 pcs. The developed system has successfully helped Toko Saing in managing stock more efficiently and providing data-based marketing recommendations. This study proves that Random Forest is effectively applied to stock prediction problems in retail goods
