Table 5.

Aggregate mean normalised performance of the forecasting models (NCB, SIR, SEIR, SIS, ARIMA, ETS) when predicting Argentina, Brazil, China, France, Germany, India, Iran, Italy, Japan, Korea, South, Spain, United Kingdom, US time series (Training phase). The rank of each model is in parentheses, in the last line.

series	Worst	Best	NCB	SIR	SEIR	SIS	ARIMA	ETS
Argentina	SEIR	NCB	0.06	0.11	0.88	0.2	0.26	0.51
Brazil	SEIR	NCB	0.01	0.04	0.89	0.13	0.2	0.18
China	SEIR	SIR	0.06	0.02	0.96	0.38	0.29	0.28
France	ARIMA	NCB	0.02	0.11	0.72	0.09	0.83	0.83
Germany	SEIR	NCB	0.01	0.13	0.96	0.14	0.22	0.22
India	SEIR	NCB	0.02	0.17	0.99	0.2	0.32	0.23
Iran	SIR	ARIMA, ETS	0.16	0.94	0.41	0.93	0.05	0.05
Italy	SEIR	NCB	0.04	0.28	1	0.36	0.06	0.06
Japan	SEIR	SIS	0.11	0.26	0.89	0.07	0.45	0.44
Korea, South	SEIR	SIR	0.22	0.04	1	0.16	0.06	0.06
Spain	SEIR	NCB	0	0.15	1	0.15	0.11	0.12
United Kingdom	SEIR	NCB	0.01	0.17	0.96	0.18	0.17	0.13
US	SEIR	ETS	0.05	0.21	0.93	0.23	0.09	0.02
APM (rank)	SEIR	NCB	0.059 (1)	0.202 (2)	0.892 (6)	0.248 (5)	0.239 (3)	0.241 (4)