A novel intelligent approach for flight delay prediction

Journal of Big Data

Table 2 Accuracy of models training

Models	Traditional training model			Proposed FDPP-ML			Error reduction percent by FDPP-ML
Models	MAE	MSE	RMSE	MAE	MSE	RMSE	MAE (%)	MSE (%)	RMSE (%)
CATR	23.37	2593	50.92	13.36	1321	36.34	43	49	29
GBR	23.74	2641	51.39	14.47	1485	38.54	39	44	25
GRU	23.68	2647	51.45	14.49	1493	38.64	39	44	25
LGR	23.52	2616	51.14	13.65	1385	37.21	42	47	27
LR	24.03	2672	51.69	17.76	1863	43.17	26	30	16
LSTM	24.40	2660	51.58	15.49	1568	39.60	37	41	23
RFR	18.28	1669	40.86	9.21	716	26.76	50	57	34
RNN	23.94	2667	51.64	14.99	1488	38.57	37	44	25
Stacking	21.85	2429	49.28	10.71	1026	32.04	51	58	35
Voting	21.57	2244	47.37	12.18	1140	33.77	44	49	29
Average error reduction percent by FDPP-ML							41	46	27