Skip to main content
. 2020 Aug 28;8:159915–159930. doi: 10.1109/ACCESS.2020.3019989

FIGURE 1.

FIGURE 1.

A schematic representation of the entire pipeline of the DeepCOVIDNet model which takes as input raw features and outputs a probability distribution to predict the range of the rise in infected cases in a given county on a given date in the future. Inline graphic correspond to number of features, Inline graphic refers to the number of time steps (experimentally set to 13), Inline graphic refers to the number of counties (set to 3146), Inline graphic refers to the dimension of each embedding, and Inline graphic refers to any non-linearity. More explanation related to the tensor shapes is presented in Section IV-A.