Data Mining & Transfer Learning for Modeling COVID


This transdisciplinary research unites infectious disease epidemiology, data science, and modeling.It will generate additional datasets that help characterize the COVID-19 pandemic system more comprehensively across socio-cultural backgrounds. The novel multivariate data-driven deep learning models will benefit from these rich set of data, reduce uncertainty, and provide higher predictability for more informed decision making during the pandemic. This project echoes the vision of MIDAS to improve the understanding of infectious disease dynamics. The study builds on my current research agenda on:1) data mining, translation, and archiving of neglected public resources with respect to COVID-19;2) data mining and annotation of social media data to evaluate public perception, attitude towards COVID-19 especially non-pharmaceutical interventions; 3) cross-cultural modeling of COVID-19 pandemic with data-driven multivariate deep learning methods. In addition to these efforts, my collaborative research on clinical stages of COVID-19 patients will also help me better understand the clinical aspect of this unprecedented disease system.


Funding Source

Midas Coordination Center Urgent Grant Program - Supplemental

Project Period