Table 2.

Forecast error for Syrian and English cities. The table shows the mean square error ($MSE$) between the forecast provided by each model ${\hat{x}}_i(t)$ and the true value $x_i(t)$. The column $\mathrm{\Delta }$ stands for the improvement of the prediction errors of the VAR model in respect to the AR model.

	Preceding	Average	AR	VAR	$\mathrm{\Delta }$($\%$)
Damascus	0.717	0.552	0.500	0.478	$+4.4$
Aleppo	1.383	1.064	1.016	1.006	$+1.0$
Idlib	1.295	0.823	0.787	0.791	$-0.5$
Daraa	0.901	0.638	0.635	0.570	$+10.2$
Homs	1.173	0.711	0.831	0.732	$+11.9$
London	0.017	0.017	0.011	0.011	$0$
Birmingham	0.088	0.059	0.052	0.051	$+1.9$
Leeds	0.110	0.066	0.061	0.060	$+1.6$
Liverpool	0.187	0.101	0.101	0.100	$+1.0$
Manchester	0.194	0.113	0.108	0.108	$0$