With the generous support of the UT Dallas Center for Disease Dynamics and Statistics, which was organized by the Department of Mathematical Sciences immediately following the outbreak of the pandemic under the auspices of the Office of the Vice President of Research, students in Dr. Qiwei Li’s class developed several real-time web apps to forecast the spread of the coronavirus with these two aims: to inform the public and local decision-makers in short order. Those analytical tools are based on several novel Bayesian methodologies.
A Bayesian model is a statistical model where probability is used to represent all uncertainty within the model, both the uncertainty regarding the output but also the uncertainty regarding the input [parameters] to the model.
“We developed a unified Bayesian statistical model to integrate different epidemiological modeling approaches for yielding better short-term forecasting of COVID-19 cases. We also developed the first COVID-19 change-point detection model to evaluate public health interventions and identify latent events associated with spreading rates,” Li said.
Random-effects change-point models are formulated for longitudinal data obtained from cognitive tests. Li said the conditional distribution of the response variable in a change point model is often assumed to be normal even if the response variable is discrete and shows ceiling effects.
“Given the observed COVID-19 case numbers over time, the model aims to detect the specific time when spreading rates change significantly due to many reasons, interventions, or virus mutation, for example.
“We can view that the model aims to divide the entire period into several time segments, each of which share the same spreading rates,” Li said.
Thanks to the valuable suggestions from Drs. Vladimir Dragovic, Pankaj Choudhary, Swati Biswas, and Min Chen, no time was lost in moving this project forward, Li said.
“We have presented the related work at multiple national and local conferences and published two papers in the Giga Science and the Journal of Data Sciences at the beginning of 2021. We will continue to pursue the project aggressively till the end of the pandemic,” Li said.
Data science skills are currently in high demand. Whether you are looking for a certificate or a degree program, the Department of Mathematical Sciences is the right place for you to learn these skills.
Stories for the newsletter, Math Matters, are produced by faculty and staff of the Department of Mathematical Sciences.
Help us leave the planet a better place for future generations. Your support for the School of Natural Sciences and Mathematics funds scientific discoveries with real-world applications, student and faculty recruitment, and academic scholarships.