Interval Short-Term Traffic Flow Prediction Method Based on CEEMDAN-SE Nosie Reduction and LSTM Optimized by GWO

Table 3

The prediction results of GWO-LSTM.

Features	Iterations	LSTM		GWO-LSTM
		Train	Test	Train	Test
1/2	100	0.0319	0.0291	0.0210	0.0197
1/2	200	0.0305	0.0282	0.0153	0.0125
1/2	300	0.0269	0.0253	0.0109	0.0068
3/4	100	0.0286	0.0253	0.0219	0.0205
3/4	200	0.0260	0.0244	0.0175	0.0186
3/4	300	0.0245	0.0226	0.0130	0.0152
2/3/4	100	0.0218	0.0188	0.0090	0.0107
2/3/4	200	0.0185	0.0169	0.0081	0.0069
2/3/4	300	0.0160	0.0120	0.0075	0.0048