COVID 19 Tracking and Modeling – UROP Spring Symposium 2021

COVID 19 Tracking and Modeling

Yitao Huang


Pronouns: He/His

Research Mentor(s): Thomas Schwarz, Associate Professor of Physics
Research Mentor School/College/Department: Physics, College of Literature, Science, and the Arts
Presentation Date: Thursday, April 22, 2021
Session: Session 5 (3pm-3:50pm)
Breakout Room: Room 20
Presenter: 3

Event Link


To build the model for COVID-19 confirmed and death cases, our group used several approaches to predict the data. We first build the model by Ridge Regression and it turns out to be pretty good on ordinary input. However, for states like MI which have a large fluctuation in data, Ridge Regression performs poorly. Then, we start to use the Neural Network approach and it turns out to be better at unordinary data; however, it has the problem of overfitting. After discussion, we add the social mobility data into account and it greatly reduces the error. Currently, we are working on the approach to predict state-level and add all the models up to predict the whole US.

Authors: Thomas Schwarz, Yitao Huang
Research Method: Computer Programming

lsa logoum logo