Environmental Systems Research

Table 3 Statistical comparison of forecasting models

From: Public warning systems for forecasting ambient ozone pollution in Kuwait

		Persistence	Time Series Modeling	Fuzzy Modeling
*AL-JAHRA*
*Dataset1 (Episodes = 2)*	RMSE	0.999	0.855	0.865
	IA	0.701	0.704	0.720
	Correct Prediction	0	0	1
	False Positive	2	2	0
*Dataset2 (Episodes = 9)*	RMSE	1.149	0.983	0.966
	IA	0.574	0.554	0.585
	Correct Prediction	1	1	1
	False Positive	8	6	3
*Dataset3 (Episodes = 3)*	RMSE	1.244	1.040	1.035
	IA	0.479	0.500	0.529
	Correct Prediction	0	0	0
	False Positive	3	3	2
*Dataset4 (Episodes = 15)*	RMSE	1.152	0.991	0.926
	IA	0.568	0.529	0.655
	Correct Prediction	2	2	3
	False Positive	11	7	6
*AL-MANSOURIYA*
*Dataset1 (Episodes = 25)*	RMSE	0.793	0.885	0.785
	IA	0.822	0.414	0.758
	Correct Prediction	15	1	1
	False Positive	9	5	0
*Dataset2 (Episodes = 0)*	RMSE	0.941	0.9154	0.861
	IA	0.741	0.3891	0.765
	Correct Prediction	0	0	0
	False Positive	0	0	0
*AL-RIGGA*
*Dataset1 (Episodes = 12)*	RMSE	0.991	0.802	0.801
	IA	0.693	0.701	0.712
	Correct Prediction	5	2	1
	False Positive	7	5	2
*Dataset2 (Episodes = 0)*	RMSE	1.090	0.862	0.850
	IA	0.643	0.574	0.662
	Correct prediction	0	0	0
	False Positive	0	0	0

Back to article page