COVID-19 Prediction

Team Members

Mingchao Lian (Team Leader)

Jiajian Lin

Jason Lai

Zijian Zhao

Yuanyuan Xiong

Project Structure

The entirety of this project is done within a Jupyter Notebook. There are two models, a recurrent neural-network (RNN) implemented within RNN.ipynb and an auto regressive integrated moving average (ARIMA) implemented within ARIMA.ipynb. We would like to be graded with our ARIMA model, as our RNN model does not produce accurate results. A more detailed explanation will be included in our final project report.

All data is included in the data directory, where data/round1/ contains all the necessary training data for round 1, data/round2/ contains all the necessary training data for round 2, and data/submission/ contains all the submission data generated from our models.

The requirements.txt file details the packages needed for this project.

Setting Up Local Environment

This project uses a Python virtual environment for development and testing. To set up a virtual environment for this project, run the following commands on your local shell.

python3 -m venv ./venv
source venv/bin/activate
pip install -r requirements.txt

Train and Test the Models

1. First, launch Jupyter Notebook with the following command:

jupyter notebook

1. Open the ARIMA_final_submission.ipynb notebook and execute all the cells in ARIMA_final_submission.ipynb.

   1. You should not have to change anything. All you have to do is run each cell one after the other from top to bottom. Prediction results generated by the models can be found under data/submission/.

   2. Our submission file for round1 is data/submission/team4_round1.csv

   3. Our submission file for round2 is data/submission/team4_round2.csv

Shutting Down Local Environment

Close all Jupyter Notebooks, then terminate the process in your shell. Run the following command in your shell to deactivate the virtual environment for this project.

deactivate