Covid-19 Daily Death Prediction Models

The code in this repository was developed alongside the writing of the following research paper:

This repository contains a linear regression model and a decision tree regression model that will predict the number of deaths that a state can expect on a given date based on CDC data and Tweets regarding Covid-19.

Model Creation and Testing

To create and view the results of our model follow the steps below.

1. Go to the /data folder and unzip the file in their called stateTwitter.zip. This should create a .csv file in the data folder called stateDate.csv. The stateTwitter.zip file is a zipped version of the training data for the models we generate.

2. Run DecisionTreeModelGenerator.py or LinRegModelGenerator.py. In the terminal, you will see the output of 10-fold cross validation run on the model with a variety of configuration parameters. For more information about how the models are generated and tested, please view the comments in the python files mentioned earlier.

3. For more information on how the stateTwitter.zip dataset was created or how to extend it, please visit the /preprocessing folder.