Estimation of
Parameters for the SEIQRD Model
Bhatt Himanshu
Res. J. Biotech.; Vol. 20(8); 232-236;
doi: https://doi.org/10.25303/208rjbt2320236; (2025)
Abstract
Accurately quantifying uncertainty in data-driven mechanistic models is crucial
for public health applications. COVID-19, a complex disease with significant global
health and economic ramifications, underscores this necessity. The pandemic's widespread
infections, mortality and economic disruptions highlight the critical importance
of understanding viral behavior and generating reliable short- and long-term forecasts
of daily new cases. Machine learning and mathematical models are actively deployed
in this effort.
To guide disease management strategies, researchers have employed diverse mathematical
models to analyze the intricate transmission dynamics of COVID-19 under varying
assumptions. This study presents the application of a six-compartment SEIQRD epidemiological
model for estimating active COVID-19 cases and deaths. Parameter estimation is achieved
through Approximate Bayesian Computation (ABC), leveraging the M-nearest neighbour
Sequential Monte Carlo ABC method which delivers the estimated parameter values.
