Home / Articles
Bayesian Analysis of Covid-19 mortality rate with Inverse Pareto distribution
Among different lethal viruses, in recent times the covid-19 was a worldwide pandemic, that the human race of the modern era of 21st century faced and it shut down the whole world for an instance. It not only effects the human race but it effects all the aspects for human life economic and social.
About 577M of cases and more than 6.4M deaths had been registered during the pandemic up until now. In Pakistan, up until now there are 1.55M cases and more than 30487 deaths had been recorded. This is crucial to apprehend the shape and flows of a disease whilst it input in a community. Moreover, to get rid of maximum loss we ought to recognize the predicted numbers of infectious patients, deaths, and the risk factors related to it. To estimate the predicted number of deaths from coronavirus, we use the Inverse Pareto distribution below Bayesian paradigm. The posterior distributions are derived assuming the non-informative priors (Uniform and Jeffery). The Bayesian estimation is done with each symmetrical and asymmetrical loss functions i.e. (Squared error, Quadratic error, Precautionary error, and weighted error).
Keywords: Inverse Pareto distribution, Uniform and Jeffery priors, loss functions, real data analysis.
- The fundamental motive of this study is to check the overall performance of every estimator under the uniform and Jeffery priors by using the real data in addition to simulated records.
In both cases under a uniform and Jeffery priors, the quadratic error loss function leads to the better estimation of the expected number of deaths due to coronavirus.