A'Yunin Sofro, Romy Ramadan Elhakim, Khusnia Nurul Khikmah
The coronavirus (COVID-19) spread rapidly across Indonesia. A month after it began in Indonesia, as many as 1790 instances had occurred. If the amounts of instances that will occur the next day can be forecast, the government can either prevent or prepare for them. This post will go over how COVID-19 data in Indonesia was analysed to forecast the number of instances the next day. This analysis employs time series estimation and data counting. The observation driven model (ODM) is one of two comprehensive approximations that may be used in data analysis. This class is also split into two parts: parametric and non-parametric. The non-parametric model employed is integer-valued autoregressive (INAR), while the parametric model is autoregressive conditional Poisson (ACP). The best model made will be picked from the two. The selection is based on each model's Akaike information criteria (AIC) value, which was generated from the findings of the ACP(1,1) model outperforming INAR(4) for the prediction of COVID-19 data in Indonesia. The ACP(1,1) model estimated that there will be up to 1661 COVID-19 positive cases in Indonesia on August 13, 2020. The total number of expected instances is 132.379. This forecast has an average absolute error (MAE) of 103.7. © 2025 Author(s).
Actuarial Sciences Department, Universitas Negeri Surabaya, Surabaya, Indonesia; Mathematics Department, Universitas Negeri Surabaya, Surabaya, Indonesia
