<![CDATA[Manual stock management at Prolab Medika Clinical Laboratory often causes delays in reporting and potential errors in data recording. Accurate stock prediction is key to avoiding shortages or excess inventory that can disrupt laboratory operations. This research aims to develop a web-based stock prediction system using machine learning to improve inventory management efficiency. The machine learning method applied is Multiple Linear Regression with variables of incoming stock, remaining stock, and outgoing stock obtained from laboratory historical data. The research results show that the system is able to predict stock requirements with a good level of accuracy based on evaluation using Mean Absolute Percentage Error (MAPE), Mean Squared Error (MSE), and R² Score, and provides real-time reports that facilitate the head of logistics in decision making.]]>
Copyrights © 2025
