EpidRLearn: Learning Intervention Strategies for Epidemics with Reinforcement Learning

Epidemics of infectious diseases can pose a serious threat to public health and the global economy. Despite scientific advances, containment and mitigation of infectious diseases remain a challenging task. In this paper, we investigate the potential of reinforcement learning as a decision making tool for epidemic control by constructing a deep Reinforcement Learning simulator, called EpidRLearn, composed of a contact-based, age-structured extension of the SEIR compartmental model, referred to as C-SEIR. We evaluate EpidRLearn by comparing the learned policies to two deterministic policy baselines. We further assess our reward function by integrating an alternative reward into our deep RL model. The experimental evaluation indicates that deep reinforcement learning has the potential of learning useful policies under complex epidemiological models and large state spaces for the mitigation of infectious diseases, with a focus on COVID-19. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Sentiment Nowcasting During the COVID-19 Pandemic

In response to the COVID-19 pandemic, governments around the world are taking a wide range of measures. Previous research on COVID-19 has focused on disease spreading, epidemic curves, measures to contain it, confirmed cases, and deaths. In this work, we sought to explore another essential aspect of this pandemic, how do people feel and react to this reality and the impact on their emotional well-being. For that reason, we propose using epidemic indicators and government policy responses to estimate the sentiment, as this is expressed on Twitter. We develop a nowcasting approach that exploits the time series of epidemic indicators and the measures taken in response to the COVID-19 outbreak in the United States of America to predict the public sentiment at a daily frequency. Using machine learning models, we improve the short-term forecasting accuracy of autoregressive models, revealing the value of incorporating the additional data in the predictive models. We then provide explanations to the indicators and measures that drive the predictions for specific dates. Our work provides evidence that data about the way COVID-19 evolves along with the measures taken in response to the COVID-19 outbreak can be used effectively to improve sentiment nowcasting and gain insights into people’s current emotional state. © 2021, Springer Nature Switzerland AG.