Time series prediction of under-five mortality rates for Nigeria: comparative analysis of artificial neural networks, Holt-Winters exponential smoothing and autoregressive integrated moving average models

Table 2 Deming regression for comparing GMDH-type ANN, ARIMA and Holt-Winters models

Reference: Observed historical U5MR	Proportional difference (slope)			Systematic difference (intercept)
Reference: Observed historical U5MR	β₁ (SE)	95% LCL, UCL	P-value	β₀(SE)	95% LCL, UCL	P-value
GMDH-type ANN	1.000 (0.0004)	0.999, 1.001	< 0.001	0.004 (0.058)	−0.113, 0.122	0.940
ARIMA	1.000 (0.001)	0.998, 1.002	< 0.001	0.027 (0.160)	−0.293, 0.348	0.865
Holt-Winters	1.000 (0.013)	0.969, 1.023	< 0.001	0.890 (2.349)	−3.822, 5.602	0.706

LCL Lower Confidence Limit, UCL Upper Confidence Limit, SE Jack-knife standard errors

ISSN: 1471-2288