Grand Award

Second Place

Predicting Covid-19 Using machine Learning: An Integration of Wastewater and Google Search Data

Data Science and Math
Ellie Dorland

Steve Beall

This research delves into the innovative use of machine learning to forecast Covid-19 trends, utilizing data from wastewater surveillance and Google search patterns. By leveraging Facebook Prophet, a sophisticated machine learning tool, this research analyzes time-series data to predict Covid-19 trends in Tucson, Arizona. The results, characterized by varying values of Root Mean Squared Error (RMSE), demonstrate the model's capability to forecast short-term trends while acknowledging challenges in long-term predictions. While there is inaccuracy in the predictions as they grow further from the known data, the model is able to track and predict the directional trend of the prevalence of Covid-19 along with providing fairly accurate predictions for the first seven days.

Project presentation

View Project Presentation file

Research paper

View Research Paper file