COMPARISON OF RANDOM FOREST AND DECOMPOSITION METHODS IN THE PREDICTION OF CENTRAL OXYGEN SUPPLY AT dr. RADEN SOEDJONO SELONG HOSPITAL

Abstract

Forecasting is a method for estimating a future value using past data. In forecasting, there are several methods, including Random Forest and Decomposition. Both methods have the advantage that they do not require assumptions compared to other forecasting methods. Issues related to Covid-19 have not yet been resolved and have caused many negative effects, one of which is the scarcity of medical gases such as oxygen. In mid-2021, oxygen to scarcity occurred in Java and Bali because of the Covid-19. This scarcity affects the supply of oxygen to other central hospitals in Indonesia, one of which is RSUD dr.R.Soedjono Selong NTB. For this reason, a strategy is needed in dealing with cases of central oxygen scarcity. One strategy that can be done is to make predictions to estimate the amount of oxygen supply each week, and the methods that can be used are Random Forest and Decomposition. The purpose of this study was to compare the Random Forest and Decomposition on the prediction of oxygen supply at RSUD dr. Raden Soedjono. The results of the analysis show that Decomposition is better than Random Forest. This is because the MAPE value produced by Decomposition (25.68%) is smaller than Random Forest (52.31%).