A novel intelligent approach for flight delay prediction

Journal of Big Data

Table 6 Accuracy of core 30 airports

NO	Airports IATA Code	Traditional training model			Proposed FDPP-ML			Error reduction percent by FDPP-ML
NO	Airports IATA Code	MAE	MSE	RMSE	MAE	MSE	RMSE	MAE (%)	MSE (%)	RMSE (%)
1	ATL	17	1074	33	10	502	22	44	53	32
2	BOS	32	3485	59	22	2503	50	31	28	15
3	BWI	25	2528	50	16	919	30	38	64	40
4	CLT	17	841	29	11	638	25	34	24	13
5	DCA	24	1636	40	13	734	27	48	55	33
6	DEN	28	2586	51	17	1079	33	40	58	35
7	DFW	24	3441	59	15	3055	55	40	11	6
8	DTW	27	3317	58	17	2640	51	36	20	11
9	EWR	57	21885	148	43	12873	113	25	41	23
10	FLL	15	394	20	12	394	20	21	0	0
11	HNL	17	689	26	14	449	21	18	35	19
12	IAD	32	4800	69	21	3620	60	34	25	13
13	IAH	23	1466	38	14	652	26	41	55	33
14	JFK	21	1157	34	16	932	31	22	19	10
15	LAS	18	1263	36	13	965	31	30	24	13
16	LAX	19	1300	36	14	487	22	29	63	39
17	LGA	22	956	31	12	453	21	47	53	31
18	MCO	18	707	27	12	406	20	33	43	24
19	MDW	17	922	30	10	221	15	40	76	51
20	MEM	29	3087	56	13	553	24	56	82	58
21	MIA	26	7798	88	21	7568	87	19	3	1
22	MSP	66	15660	125	40	7413	86	39	53	31
23	ORD	42	7405	86	29	4536	67	33	39	22
24	PHL	31	2760	53	24	2314	48	22	16	8
25	PHX	18	836	29	11	529	23	39	37	20
26	SAN	18	1294	36	10	228	15	45	82	58
27	SEA	20	834	29	13	361	19	36	57	34
28	SFO	21	774	28	13	390	20	39	50	29
29	SLC	14	338	18	9	124	11	37	63	39
30	TPA	20	1580	40	15	1192	35	25	25	13
Average error reduction percent by FDPP-ML								35	42	25