From Prediction to Prescription: AI-Based Optimization of Non-Pharmaceutical Interventions for the COVID-19 Pandemic.

Several models have been developed to predict how the COVID-19 pandemicspreads, and how it could be contained with non-pharmaceutical interventions(NPIs) such as social distancing restrictions and school and business closures.This paper demonstrates how evolutionary AI could be used to facilitate thenext step, i.e. determining most effective intervention strategiesautomatically. Through evolutionary surrogate-assisted prescription (ESP), itis possible to generate a large number of candidate strategies and evaluatethem with predictive models. In principle, strategies can be customized fordifferent countries and locales, and balance the need to contain the pandemicand the need to minimize their economic impact. While still limited byavailable data, early experiments suggest that workplace and schoolrestrictions are the most important and need to be designed carefully. It alsodemonstrates that results of lifting restrictions can be unreliable, andsuggests creative ways in which restrictions can be implemented softly, e.g. byalternating them over time. As more data becomes available, the approach can beincreasingly useful in dealing with COVID-19 as well as possible futurepandemics.