Open AccessPrediksi Akhir Pandemi COVID-19 di Indonesia dengan Simulasi Berbasis Model Pertumbuhan Parametrik
Author(s) -
Fransiscus Rian Pratikto
Publication year - 2020
Publication title -
jurnal rekayasa sistem industri/jurnal rekayasa sistem industri
Language(s) - English
Resource type - Journals
eISSN - 2339-1499
pISSN - 0216-1036
DOI - 10.26593/jrsi.v9i2.4018.63-68
Subject(s) - gompertz function , pandemic , confidence interval , statistics , covid-19 , population , mathematics , parametric statistics , demography , medicine , infectious disease (medical specialty) , disease , pathology , sociology
This research aims to predict the end of the COVID-19 pandemic in Indonesia based on parametric growth models. The models are chosen by considering their fitness with the data of Taiwan which is believed to have passed over the peak of the pandemic and have gone through all phases in the growth curves. The models are parameterized using the nonlinear least squares method. The deviation and confidence interval of each parameter is estimated using the k-fold cross-validation and the bootstrap techniques. Using the total cases per million population data from March 2 to June 18, 2020, it was found that two growth models fit the data, i.e. logistic and modified Gompertz, where the latter performs better. Using the information about the deviation of each model parameter, a simulation model is developed to predict the time at which the total cases curve starts to flatten, which is an indication of the end of the pandemic. It was found with 95% confidence level that based on the modified Gompertz model the pandemic will end somewhere between March 9 – September 7, 2021 with total cases per million of 206 - 555. Meanwhile, based on the logistic growth model, the end of the pandemic is between August 28 – September 23, 2020 with total cases per million of 180 - 375. This model can be extended by making comparative scenario with Taiwan based on measures that represent the quality of the pandemic mitigation such as test ratio and the intensity of social restriction.