Electronic Theses and Dissertation

Penyebaran virus covid-19 yang melanda dunia terjadi sejak akhir desember 2019 sangat berdampak di berbagai bidang, seperti pendidikan, kesehatan, ekonomi, dan bidang lainnya. perkembangan kasus harian covid-19 di indonesia mengalami pergerakan naik dan turun di waktu-waktu tertentu yang membuat pemerintah melakukan kebijakan seperti lockdown demi memutus rantai penularan virus covid-19 tersebut. penelitian ini bertujuan untuk melihat perkembangan kasus harian positif covid-19 periode 21 desember 2021 sampai 31 januari 2022 menggunakan peramalan deret waktu dengan metode autoregressive fractionally integrated moving average (arfima) yang merupakan pengembangan dari metode autoregressive integrated moving average (arima). peramalan menggunakan arfima memiliki syarat bahwa data harus mengandung pola long memory yaitu keterkaitan data dalam jangka panjang. penelitian ini menggunakan data kasus harian positif covid-19 di indonesia periode 13 maret 2020 sampai 20 desember 2021 sebanyak 648 observasi. hasil dari penelitian ini diperoleh model arfima terbaik adalah arfima (1, [-0,4830], 8) dengan nilai aic sebesar 1489,7140, bic sebesar 1543,4010, mae sebesar 16,1513, rmse sebesar 19,8929, dan mape sebesar 20,7837%. peramalan kasus harian positif covid-19 di indonesia periode 21 desember 2021 sampai 31 januari 2022 mengalami kenaikan dan berpola mirip dengan pola setengah data aktual. kata kunci: covid-19, deret waktu, arfima, model, peramalan

The spread of the Covid-19 virus that has hit the world since the end of December 2019 has greatly impacted various fields, such as education, health, economy, and other fields. The development of daily Covid-19 cases in Indonesia experiences up and down movements at certain times, which makes the government carry out policies such as lockdown to break the chain of transmission of the Covid-19 virus. This study aims to see the development of daily positive Covid-19 cases for the period 21 December 2021 to 31 January 2022 using time series forecasting with the Autoregressive Fractionally Integrated Moving Average (ARFIMA) method which is the development of the Autoregressive Integrated Moving Average (ARIMA) method. Forecasting using ARFIMA has a condition that the data must contain a long memory pattern, namely the relationship of data in the long term. This study uses data on daily positive cases of Covid-19 in Indonesia for the period 13 March 2020 to 20 December 2021 as many as 648 observations. The results of this study obtained that the best ARFIMA model is ARFIMA (1, [-0.4830], 8) with an AIC value of 1489.7140, BIC of 1543.4010, MAE of 16.1513, RMSE of 19.8929, and MAPE of 20.7837%. Forecasting daily positive cases of Covid-19 in Indonesia for the period from December 21, 2021 to January 31, 2022, has increased and has a similar pattern to the pattern of half the actual data. Keywords: Covid-19, time series, ARFIMA, model, forecasting