Forecasting COVID-19 Cases in the Philippines Using Various Mathematical Models

  • Edd Francis O. Felix University of the Philippines Los Baños and University of the Philippines System
  • Monica C. Torres University of the Philippines Los Baños and University of the Philippines System
  • Christian Alvin H. Buhat Department of Mathematics, University of Houston, Texas
  • Ben Paul B. Dela Cruz University of the Philippines Los Baños and University of the Philippines System
  • Eleanor B. Gemida University of the Philippines Los Baños and University of the Philippines System
  • Jonathan B. Mamplata University of the Philippines Los Baños and University of the Philippines System

Abstract

Due to the rapid increase of COVID-19 infection cases in many countries such as the Philippines, efforts in forecasting daily infections have been made to better manage the pandemic and respond effectively. In this study, we considered the cumulative COVID-19 infection cases in the Philippines from 6 March 2020 to 31 July 2020, and forecasted the cases from 1–15 August 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 compared the results to the actual data using traditional error metrics. Our results showed that the ARIMA (1,2,1) model had the closest forecast values to the actual data. Policymakers can use this result in determining which forecast method to use for their community to have data-based information for the preparation of their personnel and facilities.

Keywords: forecasting · epidemics · moving average · exponential smoothing · ARIMA · Ornstein-Uhlenbeck · SEIR · random forest

Published
2023-04-28
Section
Interdisciplinary Studies on Health (in partnership with AMDABiDSS-Health)