Model Peramalan Kasus COVID-19 di Indonesia

Arindo, Muhammad Iqbal Prathama (2022) Model Peramalan Kasus COVID-19 di Indonesia. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 09211950094005-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2024. Download (3MB) | Request a copy

Abstract

COVID-19 membawa dampak negatif pada sektor kesehatan, pendidikan dan ekonomi. Tercatat, akibat adanya COVID-19 ini economy growth Indonesia berada pada angka -2,07% pada tahun 2020. Selain itu, penelitian yang dilakukan oleh World Bank menemukan fakta bahwa efektivitas pembelajaran daring di Indonesia hanya 40%. Mengingat ekonomi, kesehatan dan pendidikan adalah aspek penting dalam sebuah negara, maka diperlukan perencanaan dan perumusan kebijakan yang sesuai dengan kondisi terkini. Pengambilan keputusan terkait perencanaan dan kebijakan tentu harus didasari oleh perkembangan kasus COVID-19 di sebuah negara. Oleh karenanya sangat penting untuk mengetahui kondisi COVID-19 di sebuah negara, setidaknya negara mampu mengambil langkah preventif untuk menekan penyebaran virus antar manusia.
Dalam penelitian telah dilakukan peramalan terkait kondisi COVID-19 di Indonesia dengan menggunakan beberapa metode yakni exponential smoothing, SIRD (suspected, infected, recovered and death) model, dan ARIMA. Hasil penelitian menunjukkan bahwa mayoritas variabel cocok menggunakan double exponential smoothing untuk melakukan peramalan karena memiliki error terkecil dibandingkan simple exponential smoothing dan ARIMA.
SIRD juga bisa menjadi salah satu alternatif solusi apabila ingin melakukan peramalan sekaligus, tanpa memisah masing-masing variabel. Hal ini dikarenakan error untuk jangka waktu peramalan 77 hari dan 46 hari tidak berbeda jauh hasilnya dengan metode double exponential smoothing karena mayoritas berada pada kisaran <5% sedangkan untuk jangka waktu peramalan 31 hari memiliki hasil yang tidak jauh berbeda dengan metode ARIMA karena error yang diberikan mayoritas berada pada kisaran <20%.
================================================================================================
COVID-19 has had a negative impact on the health, education and economic sectors. It was noted, due to the COVID-19, Indonesia's economic growth was at -2.07% in 2020. In addition, research conducted by the World Bank found the fact that the effectiveness of learning in Indonesia was only 40%.Given that the economy, health and education are important aspects of a country, it is necessary to plan and formulate policies that are in accordance with current conditions. Decision-making related to planning and policy must be based on the current development of COVID-19 cases in a country. Therefore, it is very important to know the condition of COVID-19 in a country, at least the country is able to take preventive steps to suppress the spread of the virus between humans.
In this study, predictions related to the condition of COVID-19 in Indonesia have been carried out using several methods, namely exponential smoothing, SIRD (suspected, infected, recovered and death) model, and ARIMA. The results showed that the majority of variables were suitable for using double exponential smoothing for forecasting because it had the smallest error compared to simple exponential smoothing and ARIMA.
SIRD can also be an alternative solution if you want to forecast at once, without separating each variable. This is because the error for the forecasting period of 77 days and 46 days is not much different from the results with the double exponential smoothing method because the majority are in the <5% range, while the 31-day forecasting period has results that are not much different from the ARIMA method because the majority of forecasting error in the <20% range.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	COVID-19, forecasting, model selection, pemilihan metode, peramalan
Subjects:	H Social Sciences > HA Statistics > HA30.3 Time-series analysis
Divisions:	Faculty of Creative Design and Digital Business (CREABIZ) > Technology Management > 61101-(S2) Master Thesis
Depositing User:	Muhammad Iqbal Prathama Arindo
Date Deposited:	10 Feb 2022 00:49
Last Modified:	10 Feb 2022 00:49
URI:	http://repository.its.ac.id/id/eprint/93478

Actions (login required)

View Item