Predicts incoming waves in COVID-19 cases at the county level in the U.S. Built for UniHacks 2021.

Using the NYT COVID data, the risk of an imminent COVID wave in the next 14 days is predicted. Also included is a basic webpage to allow the visualization of the predictions for each county.

An imminent COVID wave is defined as a 5% increase in the 7-day moving average of daily new cases 14 days after a given date.

This project was a submission to the UniHacks hackathon. It was selected as the best in the digital health track and also the winner of best beginner hack.

A demo of the website is available on Netlify. Additionally, a pre-trained keras model is provided on GitHub at TrainedModel.h5. To use this model for predictions rather than the one generated during training, change the model_filepath variable in predict.py, otherwise the prediction will fail.

Since the NYT data is updated every day, generating new predictions requires running the following in the repo.

python3.8 preprocess.py
python3.8 predict.py

These are what were used when developing the code. Other versions may work, but are untested and cannot be assumed to work. Other dependencies are used for the website, but those are loaded by the page since NodeJS is not used.

• Tensorflow v2.3.1 using CUDA
• Python 3.8.3
• NumPy 1.18.5

Data from The New York Times, based on reports from state and local health agencies. Available on GitHub at nytimes/covid-19-data.

Map converted from Census Bureau boundary files available on census.gov.

Copyright 2021 Timothy Laskoski. Licensed under MIT.