Table 8 Diebold-Mariano Test
From: Hybrid data decomposition-based deep learning for Bitcoin prediction and algorithm trading
| Â | LMH-BiLSTM | ||
|---|---|---|---|
One-step ahead | Two-step ahead | Three-step ahead | |
MH-BiLSTM | \({-1.677}^{*}\) | \({-1.748}^{*}\) | \({-1.773}^{*}\) |
LH-BiLSTM | \({-1.986}^{*}\) | \({-1.829}^{*}\) | \({-3.490}^{***}\) |
LM-BiLSTM | \({-2.614}^{**}\) | \({-2.762}^{**}\) | \({-2.827}^{***}\) |
H-BiLSTM | \({-3.744}^{***}\) | \({-3.501}^{***}\) | \({-5.641}^{***}\) |
M-BiLSTM | \({-4.813}^{***}\) | \({-6.749}^{***}\) | \({-5.007}^{***}\) |
L-BiLSTM | \({-4.150}^{***}\) | \({-7.225}^{***}\) | \({-7.912}^{***}\) |
BiLSTM | \({-4.509}^{***}\) | \({-6.340}^{***}\) | \({-6.072}^{***}\) |
LSTM | \({-6.021}^{***}\) | \({-10.960}^{***}\) | \({-10.704}^{***}\) |
SVR | \({-7.159}^{***}\) | \({-12.613}^{***}\) | \({-12.180}^{***}\) |
LR | \({-10.662}^{***}\) | \({-18.090}^{***}\) | \({-17.118}^{***}\) |
ARIMA | \({-6.104}^{***}\) | \({-7.241}^{***}\) | \({-7.910}^{***}\) |