Statistical Modeling on COVID-19 Infected Cases and Deaths Based on Vector Auto-Regressive Model
DOI:
https://doi.org/10.9734/bpi/aobmer/v1/7285AKeywords:
Fixed and random effect models, Panel VAR model, Cointegration test, Levin-Lin-Chu unit root test, Granger causality test, Hausman test, Wald testAbstract
The goal of this chapter is to use a panel vector auto-regressive model to simulate the dynamic correlations between the number of COVID-19-infected patients and fatalities in all districts of Kerala state, India, from January 2021 to December 2021. Modeling dynamic relationships was appropriate for the random effect panel vector auto-regressive order two models. The endogenous variable, deaths (number of deaths), is modeled to explain 62% of changes. Exogenous variable deaths (-1) are highly significant, whereas exogenous variable cases (-1) are significant at a 5% level. The endogenous variable is positively impacted by both of these external factors. The other external factors, deaths (-2) and cases (-2) are not statistically significant.
