Back

Forecasting COVID-19 cases in the Philippines using various mathematical models

Torres, M. C.; Buhat, C. A. H.; Dela Cruz, B. P. C.; Felix, E. F. O.; Gemida, E. B.; Mamplata, J. B.

2020-10-12 epidemiology
10.1101/2020.10.07.20208421 medRxiv
Show abstract

Due to the rapid increase of COVID-19 infection cases in many countries such as the Philippines, many efforts in forecasting the daily infections have been made in order to better manage the pandemic, and respond effectively. In this study, we consider the cumulative COVID-19 infection cases in the Philippines from March 6 to July 31,2020 and forecast the cases from August 1 - 15, 2020 using various mathematical models --weighted moving average, exponential smoothing, Susceptible-Exposed-Infected-Recovered (SEIR) model, Ornstein-Uhlenbeck process, Autoregressive Integrated Moving Average (ARIMA) model, and random forest. We then compare the results to the actual data using traditional error metrics. Our results show that the ARIMA(1,2,1) model has the closest forecast values to the actual data. Policymakers can use our result in determining which forecast method to use for their community in order to have a data-based information for the preparation of their personnel and facilities.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.