Exploring individuals’ effective preventive measures against epidemics through reinforcement learning

Individuals’ preventive measures,as an effective way to suppress epidemic transmission and to protect themselves from infection,have attracted much academic concern,especially during the COVID-19 pandemic.In this paper,a reinforcement learning-based model is proposed to explore individuals’ effective preventive measures against epidemics.Through extensive simulations,we find that the cost of preventive measures influences the epidemic transmission process significantly.The infection scale increases as the cost of preventive measures grows,which means that the government needs to provide preventive measures with low cost to suppress the epidemic transmission.In addition,the effective preventive measures vary from individual to individual according to the social contacts.Individuals who contact with others frequently in daily life are highly recommended to take strict preventive measures to protect themselves from infection,while those who have little social contacts do not need to take any measures considering the inevitable cost.Our research contributes to exploring the effective measures for individuals,which can provide the government and individuals useful suggestions in response to epidemics.

A multi-scale agent-based model of infectious disease transmission to assess the impact of vaccination and non-pharmaceutical interventions:The COVID-19 case

Mathematical and computational models are useful tools for virtual policy experiments on infectious disease con-trol.Most models fail to provide flexible and rapid simulation of various epidemic scenarios for policy assessment.This paper establishes a multi-scale agent-based model to investigate the infectious disease propagation between cities and within a city using the knowledge from person-to-person transmission.In the model,the contact and infection of individuals at the micro scale where an agent represents a person provide insights for the interactions of agents at the meso scale where an agent refers to hundreds of individuals.Four cities with frequent population movements in China are taken as an example and actual data on traffic patterns and demographic parameters are adopted.The scenarios for dynamic propagation of infectious disease with no external measures are compared versus the scenarios with vaccination and non-pharmaceutical interventions.The model predicts that the peak of infections will decline by 67.37%with 80%vaccination rate,compared to a drop of 89.56%when isolation and quarantine measures are also in place.The results highlight the importance of controlling the source of infection by isolation and quarantine throughout the epidemic.We also study the effect when cities implement inconsis-tent public health interventions,which is common in practical situations.Based on our results,the model can be applied to COVID-19 and other infectious diseases according to the various needs of government agencies.