Affinity Workshop: WiML Workshop 1

A Data-driven Approach to Infer Latent Dynamics of COVID-19 Transmission Model

Sujin Ahn · Minhae Kwon


Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2; COVID-19) has rapidly spread across the world since 2019. Under the circumstance of viral variants that have emerged, the importance of mathematical modeling of infectious diseases is highlighted to understand the ongoing outbreak. In this work, we propose a data-driven epidemic model based on the Markov chain including a vaccinated and isolated group. Our model uses daily reported data and fits them into our model to find underlying dynamics parameters of the Markov chain. In this work, we aim to estimate latent state values by taking advantages of the officially reported data and underlying Markov chain dynamics. We confirmed that the proposed model is able to successfully estimate all states by fitting COVID-19 data in South Korea into our model.

Chat is not available.