Comparative Analysis of Decision Tree and Random Forest Algorithms in Predicting Potential Buyers and Sellers in the Healthcare Sector

The healthcare sector plays a crucial role in ensuring the availability of appropriate and high-quality medicines for patients. This research addresses the issue of overstock and medicine expiration in private hospitals by developing a predictive model using the Decision Tree C4.5 and Random Forest algorithms. Additionally, this study also helps reduce problems related to the overstocking or dead stocking of medicines. The model aims to identify hospitals within the network that may require medication from other hospitals, thus reducing the expiration risk. The implementation results show that a Random Forest is more effective than a Decision Tree. Random Forest achieved 70% accuracy, while Decision Tree reached 61%, and it also scored an ROC AUC of 0.91, higher than Decision Tree's 0.82. This comparison highlights the advantages of Random Forest. Through a user-friendly predictive website, this research enhances coordination with hospital inventory planners, improving operational efficiency and revenue potential. In this research, the result will be made in the form of a website to identify potential sellers and potential buyers from affiliated hospitals.