The coronavirus disease COVID-19 has so far infected 1,075 people in the Philippines, as of Saturday, March 28, and is significantly impacting the Philippine economy and in many countries.
We used predictive analytics, particularly triple exponential smoothing, to forecast how many confirmed COVID-19 cases can be expected in the near future including Sunday, March 29.
Our method predicts that by March 29, the virus confirmed cases will have climbed to 1,347, or between 1,257 and 1,437 confirmed cases, based on data available up to March 5.
As of when this forecasting was published, our model had predicted COVID-19 infections to an accuracy of 98% within the 95% confidence interval. To maintain this accuracy, we have to regularly update our data taken from the Department of Health.
Table 1. Prediction Values of COVID-19 Confirmed Cases in the Philippines for March 29 to April 2, 2020
Based on data available up to March 5 where we have accumulated cases of 3 since January and eventually the the number of cases increases thereafter, our model predicts that by April 2 the number of confirmed cases nationwide will surpass 2,400. However, since these projections surpass our short-term window of accuracy, especially yesterday’s forecast, they shouldn’t be considered as reliable as the figures above.
Having said that, this model is best suited for short-term forecasting. To make accurate long-term forecasts, we’ll need more historical data and a better understanding of the variables impacting COVID-19’s spread.
Also, we have provided you other two predictive analytics models to better understand the behavior of the data. These are the AI-inspired time forecasting model and the SIR model.