ABSTRACT
COVID-19, responsible for infecting billions of people and the economy across the globe, requires a detailed study of the trend it follows to develop adequate short-term prediction models for forecasting the number of future cases. In this perspective, it is possible to develop strategic planning in the public health system to avoid deaths as well as managing patients. In this paper, forecast models comprising various artificial intelligence approaches such as support vector regression (SVR), long short term memory (LSTM), bidirectional long short term memory (Bi-LSTM) are assessed for time series prediction of confirmed cases, deaths, and recoveries in ten major countries affected due to COVID-19. The paper also reviewed a deep learning model to forecast the range of increase in COVID-19 infected cases in future days to present a novel method to compute multidimensional representations of multivariate time series and multivariate spatial time series data. The paper enables the researchers to consider a large number of heterogeneous features, such as census data, intra-county mobility, inter-county mobility, social distancing data, past growth of the infection, among others, and learn complex 78interactions between these features. To fast-track further development and experimentation, the analyzed code could be used to implement the AI in an efficient way. The paper discusses existing theories and research that provide a better understanding of the spread pattern recognition which will help to tackle any future pandemic of similar intensity. We encourage others to further develop a novel modeling paradigm for infectious disease based on GNNs and high resolution mobility data.
