Table 1.
Models and data used to predict the number of confirmed cases.
Fig 1.
Structure of DNN.
Fig 2.
Structures of LSTM (left) and GRU (right).
Fig 3.
Overall framework.
Fig 4.
Diagram for expanding sentiment dictionary.
Table 2.
Process of extracting text data polarity.
Fig 5.
Layer configuration diagram.
Fig 6.
Machine learning prediction method.
Table 3.
Prior literatures’ data period examples.
Table 4.
Example of preprocessing results.
Table 5.
Example of expanded sentiment dictionary.
Fig 7.
Method for calculating daily polarities.
Table 6.
Prediction accuracy results by case.
Table 7.
Example of extracted text polarity data.
Table 8.
Example of a table of daily polarities.
Table 9.
Model accuracy results.
Fig 8.
DNN results obtained using 14-day data with polarity excluded (left) and included (right).
Fig 9.
LSTM results obtained using 14-day data with polarity excluded (left) and included (right).
Fig 10.
GRU results obtained using 14-day data with polarity excluded (left) and included (right).
Table 10.
Comparison of prediction model results at 7-day intervals.