Medicine Stock Forecasting Using Least Square Method

Abstract

A planning activities in order to ensure drug availability according category and quantity is very necessary by health organizations. Therefore, this study, Least Square (LS) method for forecasting as a part of planning in order to guarantee the drugs availability based on drugs consumption past data have been implemented. In this study, drugs consumption data in period January - November 2017 or 197 samples datasets have been utilized. Based on the experimental results, the prediction for the next month has an average accuracy value, Mean Absolute Deviation (MAD) of 51.20 %, Mean Square Error (MSE) of 66.29 % and Mean Absolute Percentage Error (MAPE) of 10% has been obtained. In other words, the LS method could be explored as an alternative forecasting method of drugs availability. Furthermore, an accuracy increased improved by using the computational intelligence (CI) method is next research plan.