Projecting the Impact of COVID-19 on the Demand using Machine Learning.
Measuring and estimating the demand impact of any product and service is crucial for economic and energy planning and policy discussions.
The Covid-19 pandemic has caused the demand for fuel to fall almost everywhere in countries that have adopted a reduction in human activities, such as total or partial containment, in order to reduce the level of spread of the virus.
In their paper « Machine learning model to project the impact of COVID-19 on US motor gasoline demand » the authors of this study (Or, S., He, X., Ji, W. et al. Machine learning model to project the impact of COVID-19 on US motor gasoline demand. Nat Energy 5, 666-673 (2020). https://doi.org/10.1038/s41560 -020-0662-1, July 2020), incorporated pandemic projections and resulting travel activities and fuel consumption into a machine learning model to project U.S. medium-term gasoline demand and study the impact of government intervention.
The authors found that « under the baseline infection scenario, U.S. gasoline demand is slowly increasing after a rapid rebound in May 2020 and is unlikely to recover completely until October 2020. Under the baseline and pessimistic scenario, continuous locking (no reopening) could temporarily increase demand for automotive fuel, but it helps demand return to normal levels more quickly. In the optimistic infection scenario, gasoline demand will recover near the non-pandemic level by October 2020 ».
