COVID 19 Predictions at RCMS NUST
-COVID 19 Predictions at RCMS NUST
Updated on 23 June
The predictions are based on actual data from John Hopkin University Corona Virus Resource Center. Two different parameter estimation techniques are used: Time invariant parameters (TI-SIR – red) and the proposed Time dependent parameters (TD-SIR – black). The results of are shown below:
Pakistan |Total Cases: 0.33M |Total Deaths: 15.6K |Peak: 15/Jun/2020 |97% End: 30/Aug/2020
World |Total Cases: 13.2M| Total Deaths: 1.18M | Peak: 11/Jun/2020 |97% End: 09/Sep/2020
The approach is based on a modified susceptible-infected-recovered (SIR) compartmental models in which the associated unknown parameters are considered time-varying and are identified through the solution of a nonlinear constrained optimisation problem. The development of this nonlinear optimisation framework for parameter estimation and comparing the applicability of the new framework with the commonly used time-invariant SIR (TI-SIR) model on projection of COVID-19 data are the main contributions of our work. It is observed that the estimation results of the time-dependent model better suites the actual data as compared to time-independent model.
Disclaimer: The results here are strictly for research and may contain errors