A Hybrid LSTM-Based Ensemble Learning Approach for China Coastal Bulk Coal Freight Index Prediction

Table 7

DM test results for hybrid models and the benchmarks.

Data type	Tested model	Reference model
		LSTM	GBRT	RF	LSTM-RF	LSTM-GBRT
Daily CBCFI forecasting	LSTM	—
	GBRT	2.4328	—
	RF	2.4103	−1.6784	—
	LSTM-RF	2.5123	2.2341	2.1231	—
	LSTM-GBRT	2.2763	2.2910	2.8723	−1.2432	—
Weekly CBCFI forecasting	LSTM	—
	GBRT	−2.7084	—
	RF	2.2034	−1.6110	—
	LSTM-RF	2.2361	2.3414	2.6535	—
	LSTM-GBRT	2.2276	2.3012	2.5541	2.6287	—
Monthly CBCFI forecasting	LSTM	—
	GBRT	−4.5123	—
	RF	−5.2341	−1.9883	—
	LSTM-RF	−5.2011	−6.1094	−5.3312	—
	LSTM-GBRT	−5.1998	−6.2014	−6.4234	−2.6536	—

Note.The value is significant at 5%.