<![CDATA[PT Raia Dian Ananta faces challenges in managing lubricant stocks which often experience irregularities, either in the form of excess stock buildup or stock shortages during times of high demand. This research aims to design and build a web-based lubricant stock demand prediction system to optimize company inventory management. The methodology used is K-Nearest Neighbors (KNN), a machine learning algorithm that performs classification or regression based on the closeness of the distance between historical data. The main parameters used in this prediction model include the number of sales in the previous period, remaining final stock, selling price, and lubricant category. This system was developed with a web-based platform to facilitate real-time data accessibility. The results show that the KNN model with K=5 provides high accuracy (MAPE 7.84%) and the developed web-based system successfully facilitates management in making effective procurement decisions. By implementing this system, PT Raia Dian Ananta can minimize unnecessary stock storage costs and increase operational efficiency through more guaranteed product availability for customers.]]>
Copyrights © 2026
