STATISTICAL MODELLING OF COVID-19 PANDEMIC

Abstract

Coronavirus pandemic data is a family of clinical data which requires the attention of flexible model for modeling its nature. In this study, modeling and analysis based on the available data gathered from each state by the Nigeria Centre for Disease Control are carried-out on Nigerian cases of COVID-19 pandemic using some selected univariate continuous models include: Cauchy, Gumbel, Logistic, Lognormal and Weibull model to fit each data set from each case such as confirmed, discharged, deaths and total active case. Percentage of each case in the affected state are obtained and presented to ascertain the level at which each state has been affected by COVID-19. Secondary data collected from Nigeria centre for disease control site are used for the study; and the data consists 36 states including FCT Abuja. Therefore, with all indications the results shown from exploratory data analysis, goodness-of-fit criteria and statistics that Lognormal model has better fit to all the cases considered in the study than other models despite the outliers in the data sets of all cases