Hybrid-Epanechnikov Transformed Kumaraswamy Distribution: Applications to COVID-19 Mortality and Survival Analysis

Abstract

In this paper, we introduce a new flexible two-parameter unit interval univariate probability distribution called the "hybrid-Epanechnikov transformed Kumaraswamy distribution (HTKD)”. The HTKD distribution, a flexible and robust version of the Kumaraswamy distribution, was created with the hybrid Epanechnikov kernel. We performed an in-depth analysis of its statistical characteristics and determined the parameters using the maximum likelihood estimation technique. The HTKD distribution's applicability was demonstrated using three datasets: the COVID-19 survival rate of Spain, the COVID-19 death rates of Canada, and the COVID-19 mortality rates of the UK. The HTKD distribution consistently offered the best fit when compared to other distributions, including the Beta, Kumaraswamy, BurrXII, and Weibull distributions, as seen by the lowest AIC and BIC values. These findings demonstrate the promise of the HTKD distribution as a flexible and useful tool for statistical analysis in epidemiological research. The uniformity of its performance across several datasets highlights its capacity to offer precise and dependable modelling of epidemiological data.