Zero-Inflated Time Series Modelling of COVID-19 Deaths in Ghana

Table 3

Dynamic zero-inflated Poisson autoregressive (DZIPA) model and dynamic negative binomial autoregressive (DZINBA) model estimate with corresponding standard errors (s.e.) for COVID-19 deaths in Ghana.

	DZIPA		DZINBA
	Log-linear		Log-linear
Coefficients	Estimate (s.e.)	Pr (>│z│)	Estimate	Pr (>│z│)
Intercept	1.1012 (0.5469)	0.0441	0.8066 (1.2224)	0.5093
Trend	−1.3129 (3.9978)	0.7426	−0.3020 (7.4510)	0.9677
Autoregressive
AR1	0.9735 (0.0172)	<0.0001	0.9618 (0.0291)	<0.0001
Zero-inflation	0.6148	—	0.6108	—
Standard deviation	0.1908	—	0.2285	—
N	200	—	200	—
R	200	—	200	—
Number of iterations	100	—	100	—
Dispersion,	—	—	11.3647	—
Goodness of fit
AIC	629.1376	—	636.9148	—
BIC	646.8336	—	658.1624	—
TIC	722.0860	—	1023.5070	—