Vector autoregression model using behavioral data successfully forecasts COVID-19 cases: An application of big data

Résumé fourni par la source

We demonstrated the usefulness of behavior-related indicators from big data in providing timely information for surveillance using vector autoregression. Impulse response analysis and forecast error variance decomposition showed variation in the response patterns of daily number of COVID-19 cases to change in mobility and public interest at three periods during the pandemic, with the increase in the number of cases in response to innovations in public interest during the resurgence due to Omicron variant. Evaluation of forecast accuracy suggested reasonable forecasts. These emphasize that data outside the field of health can fill the gap in information from traditional surveillance.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.