Penerapan Data Mining Untuk Memprediksi Pembelian Obat di RSUD Dr. Moewardi

Hospitals are an integral part of an organization of health that functions to provide completeness, cure diseases, and prevent diseases to the community. especially drugs in the health sector contained in hospitals. Purchasing drugs that are not in accordance with the needs that exist at that time makes the hospital suffer large losses. So that the author analyzes the purchase of drugs can be in accordance with existing needs based on analysis of patterns in previous years. Using the ARIMA method which has good forecast accuracy to predict the short term. By using data mining to calculate large amounts of data from previous years, it is hoped that from this prediction calculation for drug purchases can be in accordance with the needs so as not to make large losses. From all existing drug data, the author uses 3 samples of drug data, namely Nacl, Ranitidin, and Omeprazole. The results for Nacl data the accuracy rate reached 14.29%, Ranitidin data has an accuracy rate of 2.76%, and for Omeprazole data has an accuracy rate of 12.78%. From the results of time series analysis and the application of the ARIMA method, it can be concluded that data mining can be used to accurately predict drug purchases. By understanding the pattern of drug purchases over time and building a suitable ARIMA model, hospitals can make better decisions in planning drug purchases and inventory. Thus, the application of data mining in predicting drug purchases can provide significant benefits for Dr. Moewardi Hospital.