Predicting COVID-19 Cases Using Time Series Analysis with Neural Networks

Neuro 140 - Harvard College Spring 2020

Team: Jason Lee, Seeam Noor, and Taras Holovko

Project Description:

Goal of this project was to develop & compare different NNs to model & predict COVID-19 confirmed cases. The video for the project presentation can be found here: https://www.youtube.com/watch?v=XtAontZl-IE&feature=youtu.be

Models Used:

We used 3 types of models to predict Covid-19 cases:

1. Feed Forward Neural Networks
2. Long Short-term Memory Networks (LSTM)
3. Gated Recurrent Units Neural Networks (GRU)

Running the Notebook:

The code for the entire project is in the file 'code_report_vf' The 2 folders contain all the data you'd need to run the project. Make sure you have all the required packages installed!