The Prediction of Coronavirus Cases and an Analysis of the Factors Contributing to its Spread Across Universities in the United States
Open Access
Author:
Vo, Christopher
Area of Honors:
Statistics
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Lingzhou Xue, Thesis Supervisor Le Bao, Thesis Honors Advisor
Keywords:
Coronavirus Artificial Neural Network Deep Learning Multiple Linear Regression Support Vector Regression
Abstract:
In March 2020, the coronavirus started an unprecedented global health crisis. Having to suddenly rethink their way of operating under new conditions, businesses and institutions were immediately impacted. In-person events were postponed, budgets needed to be revised, and supply chains were drastically disrupted. The pandemic continued for the rest of the spring and throughout the entire summer of 2020. As the 2020-2021 school year approached, leaders of academic universities within the United States were required to make critical decisions about how to proceed with the fall semester amid a global pandemic. For this thesis, several factors that may contribute to the number of coronavirus cases at more than 80 U.S. universities were analyzed. The number of cases specified in the dataset is from March 2020 until February 26, 2021. There are two main objectives this thesis hopes to achieve: making predictions about the number of coronavirus cases for a university given the values of the analyzed factors and determining which factors contribute the most to increasing coronavirus cases on college campuses. To achieve these two objectives, a variety of models were fitted to the dataset. In particular, the implementation of multiple linear regression, support vector regression, and deep learning was completed.
