Covid-19 Retail Footfall Trends
The University of Miami–Herbert School of Business, in partnership with Deloitte, asked several members of the MSBA program to predict changes in footfall trends during the COVID-19 pandemic for a client in the retail industry.
By the Numbers
0
%
of Fortune 10,000 companies saw supply chain disruptions from COVID-19
0
%
of companies had negative impacts on their business from COVID-19
0
%
of companies have downgraded their growth outlooks
Datasets
Google Mobility Report
COVID-19 Cases Dataset
State Enforced Masked Mandate
Madated Stay at Home Order
COVID-19 Twitter Mentions
Descriptive Variables
•If a county was in the top 30 GDPs in the US
•The region were the state is located
•Which political party had a majority in the state legislature
•The population of a state
Footfall vs Covid cases
The data was uploaded to Amazon Web Services (AWS) forecasting which was used to make retail footfall predictions
AWS was additionally beneficial because the forecast took weather into account
The best model was a DeepAR+ model
This predicted footfall by implementing a one-dimensional time series using a recurrent neural network
This model had an RSME of
0
%
If you want to view the Tableau Dashboard click the image to the right
