Comparison of Conventional Modeling Techniques with the Neural Network Autoregressive Model (NNAR): Application to COVID-19 Data

Table 3

Accuracy of different time-series models for predicting COVID-19 cumulative daily deaths.

Method	Source	RMSE	MAE
Mean	Training	38.17407	30.06568
	Testing	34.05141	31.50445
Drift	Training	25.76414	14.11472
	Testing	33.95330	31.50954
Naïve	Training	25.76424	14.10708
	Testing	29.07091	26.64583
Holt	Training	20.28812	11.70446
	Testing	123.38359	98.10785
SES	Training	20.27235	11.62327
	Testing	37.35009	34.89225
ARIMA order SES	Training	20.20818	11.58092
	Testing	36.33789	33.86672
ARIMA order (2, 2, 2)	Training	20.20818	11.58092
	Testing	36.33789	33.86672
ARIMA auto	Training	20.06109	11.54640
	Testing	22.33402	20.61141
NNAR	Training	10.36647	5.065002
	Testing	12.89895	8.009270