Table 1.

Parameter estimates for the epidemiological periods (P₁–P₆).

Time Delay	Distribution	m	Mean (day)	SD* (day)	AICC	BIC
Symptom onset to confirmation for imported cases (P1)	Weibull	16	6.76 (4.53, 9.28)	4.74 (3.05, 8.70)	95.11	95.73
Symptom onset to confirmation for local cases (P2)	Log-normal	9	2.57 (1.57, 4.23)	1.99 (0.72, 4.97)	37.51	35.90
Confirmation to discharge (P3)	Weibull	24	15.91 (14.06, 17.72)	4.66 (3.63, 6.30)	147.10	148.88
Symptom onset to discharge (P4)	Weibull	21	21.87 (19.97, 23.72)	4.29 (3.42, 6.02)	124.30	125.72
Incubation period (P5)	Log-normal	9	5.53 (3.98, 8.09)	2.96 (1.28, 6.09)	47.65	46.04
Serial interval (P6)	Log-normal	9	6.45 (4.32, 9.65)	4.16 (1.67, 8.87)	52.45	50.85

m, the number of data in a dataset. SD*, standard deviation, where 95% CI is shown in parenthesis. In-sample errors were computed by the second order Akaike information criterion (AIC_C) values and the Bayesian information criterion (BIC) for three different distributions (gamma, log-normal, and Weibull). AIC_C and BIC are defined by ${\mathrm{AIC}}_{\mathrm{C}}=2n-2log\left(L\right)+\frac{2n^2+2n}{m-n-1}$ and $\mathrm{BIC}=-2log\left(L\right)+log\left(m\right)n$, where n represents the number of parameters and L is the maximized likelihood of a fitted delay function.