Table 1 Variables collected for timeseries analysis and their sources.
From: A novel bidirectional LSTM deep learning approach for COVID-19 forecasting
Description | Frequency of refresh | Data source |
|---|---|---|
Infected covid cases | Daily | COVID tracking data from Johns Hopkins Coronavirus Resource Center |
Dead covid cases | ||
Recovered covid cases | ||
Derived data on the effective reproduction number, Rt | Daily | "Country-wise forecast model for the effective reproduction number Rt of coronavirus disease," Frontiers in Physics, vol. 8, p. 304, 2020 |
Flight data for 12 countries -United states -United Kingdom -UAE -Germany -Spain -France -Japan -Korea South -China -Brazil -Sweden -Singapore | Weekly | Flight Data from Official Airline guide (OAG) website |
Covid test data, total tests, and per thousand population | Daily | Testing data from ourworldindata.com |
Level of containment policies (international travel controls, contact tracing, facial coverings, stay-home requirements) adopted by each country across the time | Daily | Containment policies from ourworldindata.com |
Mobility data - This new dataset from Google measures visitor numbers to specific categories of location (e.g. grocery stores; parks; train stations) every day and compares this change relative to the baseline day before the pandemic outbreak. Baseline days represent a normal value for that day of the week, given as the median value over the five-week period from January 3rd to February 6th, 2020. Measuring it relative to a normal value for that day of the week is helpful because people obviously often have different routines on weekends versus weekdays | Daily | Google mobility data from ourworldindata.com |
