Prediction of flight departure delays caused by weather conditions adopting data-driven approaches

Journal of Big Data

Table 14 Results of JFK airport (accuracy from 1 to 24 h)

Algorithm	1 h	2h	3 h	4 h	5 h	6 h	7 h	8 h	9 h	10 h	11 h	12 h	13 h
DT	0.831	0.787	0.828	0.790	0.826	0.806	0.804	0.796	0.789	0.791	0.791	0.790	0.799
RF	0.882	0.843	0.877	0.850	0.869	0.853	0.854	0.843	0.846	0.848	0.845	0.842	0.846
SVM	0.801	0.650	0.782	0.638	0.728	0.701	0.698	0.643	0.662	0.678	0.671	0.665	0.638
KNN	0.803	0.712	0.792	0.722	0.772	0.756	0.738	0.724	0.731	0.739	0.736	0.724	0.730
LR	0.706	0.581	0.665	0.573	0.642	0.617	0.606	0.594	0.582	0.584	0.605	0.588	0.560
XGB	0.859	0.779	0.851	0.769	0.822	0.801	0.801	0.776	0.788	0.783	0.786	0.782	0.779
LSTM	0.848	0.852	0.842	0.829	0.828	0.826	0.817	0.814	0.809	0.812	0.804	0.803	0.800

Algorithm	14 h	15 h	16 h	17 h	18 h	19 h	20 h	21 h	22 h	23 h	24 h	Avg train (s)	Avg test (us)
DT	0.799	0.790	0.800	0.786	0.797	0.784	0.781	0.796	0.797	0.784	0.779	0.081	0.796
RF	0.844	0.844	0.840	0.836	0.844	0.832	0.841	0.839	0.839	0.846	0.837	1.809	21.503
SVM	0.646	0.659	0.642	0.620	0.650	0.616	0.653	0.651	0.628	0.610	0.618	11.877	1289.089
KNN	0.727	0.724	0.725	0.728	0.724	0.708	0.722	0.727	0.722	0.707	0.723	0.009	57.606
LR	0.577	0.565	0.582	0.575	0.577	0.559	0.603	0.608	0.569	0.557	0.562	0.128	0.531
XGB	0.774	0.779	0.778	0.772	0.783	0.771	0.786	0.779	0.769	0.767	0.778	0.194	2.655
LSTM	0.795	0.809	0.799	0.804	0.803	0.796	0.794	0.786	0.785	0.785	0.778	554.2	4.513