<![CDATA[COVID-19 (Coronavirus Disease 2019) is a new type of disease related to the same virus family as Severe Acute Respiratory Syndrome (SARS) and several strains of the common cold virus. Along with the increase of positive cases, the resources needed in handling COVID-19 cases also increase. To overcome this problem, anticipatory measures are needed so that the resources needed in handling COVID-19 such as health workers and medicines will be available before positive cases spike. In this study, the method used is the hybrid Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM) method. The ARIMA-LSTM model is built by combining the ARIMA (2,1,2) model with the LSTM model which has 4 hidden states and 1 layer. ARIMA model is used to predict the trend value from time series data while LSTM model is used to complete the ARIMA model forecasting by predicting the time series residual value. Based on testing, the ARIMA-LSTM model achieved high accuracy, especially in short-term forecasting with an error rate of 1.8 percent for forecasting cases for the next 3 days.]]>
Copyrights © 2022
