A Comparison: Prediction of Death and Infected COVID-19 Cases in Indonesia Using Time Series Smoothing and LSTM Neural Network

• Published in: Procedia computer science 2021, Vol.179, p.982-988
• Main Authors: Rasjid, Zulfany Erlisa, Setiawan, Reina, Effendi, Andy
• Format: Article
• Language: English
• Subjects: Covid-19; Long Short Term Memory Neural Network model; Prediction; Savitzky Golay Smoothing
• ISSN: 1877-0509; 1877-0509
• DOI: 10.1016/j.procs.2021.01.102

COVID-19 is a virus causing pneumonia, also known as Corona Virus Disease. The first outbreak was found in Wuhan, China, in the province of Hubei on December 2019. The objective of this paper is to predict the death and infected COVID-19 in Indonesia using Savitzky Golay Smoothing and Long Short Term Memory Neural Network model (LSTM-NN). The dataset is obtained from Humanitarian Data Exchange (HDX), containing daily information on death and infected due to COVID-19. In Indonesia, the total data collected ranges from 2 March 2020 and by 26 July 2020, with a total of 147 records. The results of these two models are compared to determine the best fitted model. The curve of LSTM-NN shows an increase in death and infected cases and the Time Series also increases, however the smoothing shows a tendency to decrease. In conclusion, LSTM-NN prediction produce better result than the Savitzky Golay Smoothing. The LSTM-NN prediction shows a distinct rise and align with the actual Time Series data.