Accepted for/Published in: Journal of Medical Internet Research
Date Submitted: Sep 13, 2020
Date Accepted: May 31, 2021
Date Submitted to PubMed: Jun 3, 2021
Real-time prediction of the daily incidence of COVID-19 in all countries and territories individually using machine learning: an infodemiology study
ABSTRACT
Background:
Advanced prediction of the daily incidence of COVID-19 helps policy making on prevention of spread which profoundly affects peoples’ livelihood. Previous studies have investigated prediction in single or several countries and territories.
Objective:
We aim to develop models for real-time prediction of COVID-19 activity in all countries/territories individually worldwide.
Methods:
Data of the previous daily incidence and Google Trends from all the individual countries/territories were collected. Random Forest Regression algorithm was used to train models to predict the new confirmed cases seven days ahead. Several methods were used to optimize the models, including clustering the countries/territories, features selection according to the importance scores, multiple-step forecasting, and upgrading models at regular intervals. The performance of the models was assessed using mean absolute error (MAE), root mean square error (RMSE), Pearson correlation coefficient and Spearman correlation coefficient.
Results:
Our final models can accurately predict the daily new confirmed cases of COVID-19 in most countries/territories. There were 198 (92.1%) countries/territories with MSE <10 and 187 (87.0%) with Pearson correlation coefficient >0.8. In a total of 215 countries/territories, the mean MAE was 5.42 (range 0.26 - 15.32), the mean RMSE 9.27 (range 1.81 – 24.40), the mean Pearson correlation coefficient 0.89 (range 0.08 - 0.99), the mean Spearman correlation coefficient 0.84 (range 0.21 – 1.00). P < .001 in most of countries/territories.
Conclusions:
Integrating the previous incidence and Google Trends data, we are able to predict the incidence of COVID-19 in most individual countries/territories accurately seven days ahead using machine learning.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.