PEMODELAN JUMLAH KASUS BARU HARIAN COVID-19 DI INDONESIA MENGGUNAKAN GAUSSIAN MIXTURE MODEL

Abstract

Penyakit COVID-19 adalah penyakit menular yang disebabkan oleh virus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Selama masa pandemi COVID-19 terjadi beberapa kali lonjakan jumlah kasus baru COVID-19 yang menunjukkan adanya kesulitan dalam mengantisipasi peningkatan penyebaran COVID-19. Artikel ini membahas pemodelan jumlah kasus baru harian COVID-19 di Indonesia dari 1 Januari 2021 sampai 31 Maret 2022 menggunakan Gaussian Mixture Model (GMM). GMM merupakan salah satu mixture model dimana setiap komponen campuran diasumsikan berdistribusikan Gaussian. GMM dikonstruksi mengggunakan beberapa komponen campuran dan parameter dari setiap GMM diestimasi menggunakan metode maximum likelihood estimation (MLE) melalui algoritma Expectation-Maximization (EM). Berdasarkan nilai Akaike Information Criteria (AIC), diperoleh GMM dengan 4 komponen merupakan model terbaik untuk pemodelan data jumlah kasus baru harian COVID-19 di Indonesia. Berdasarkan menggunakan model GMM terbaik, diperoleh probabilitas jumlah kasus baru harian COVID-19 di Indonesia kurang dari jumlah kasus harian terendah adalah 0,01, lebih dari jumlah kasus harian rata-rata adalah 0,3 dan lebih dari jumlah kasus harian tertinggi adalah 0,017.

COVID-1 is an infectious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). During the COVID-19 pandemic, there were several spikes in the number of new COVID-19 cases, which shows that there are difficulties in anticipating the increase in the spread of COVID-19. This article discusses modelling the number of daily new cases of COVID-19 in Indonesia from 1st January 2021 to March 31, 2022, using the Gaussian Mixture Model (GMM). GMM is a mixture model where each mixture component is assumed to have a Gaussian distribution. GMM is constructed using several mixed components, and the parameters of each GMM are estimated using the maximum likelihood estimation (MLE) method via the Expectation-Maximization (EM) algorithm. Based on the Akaike Information Criteria (AIC) values, it was found that GMM with 4 components is the best model for modelling data on the number of daily new cases of COVID-19 in Indonesia. Based on the best GMM model, the probability that the number of new daily COVID-19 cases in Indonesia is less than the lowest number of daily cases is 0.01, more than the average number of daily cases is 0.3, and more than the highest number of daily cases is 0.017.