Table 3 Prediction errors of tourist flow by different models at mount Lu.
From: Combined CNN-BiLSTM-Att tourism flow prediction based on VMD-MWPE decomposition reconstruction
Model | 3 days | 7 days | 15 days | ||||||
|---|---|---|---|---|---|---|---|---|---|
MAPE (%) | BRMSE (%) | R2 | MAPE (%) | BRMSE (%) | R2 | MAPE (%) | BRMSE (%) | R2 | |
ARIMA | 10.78 | 13.49 | 0.829 | 12.18 | 12.84 | 0.743 | 18.69 | 18.84 | 0.628 |
SVM | 13.89 | 15.27 | 0.761 | 15.91 | 15.15 | 0.665 | 23.74 | 25.66 | 0.513 |
DT | 3.48 | 3.76 | 0.946 | 5.53 | 6.76 | 0.891 | 13.42 | 14.21 | 0.694 |
RF | 3.96 | 4.15 | 0.937 | 5.26 | 6.07 | 0.896 | 12.63 | 13.93 | 0.729 |
DNN | 4.19 | 4.27 | 0.932 | 4.86 | 5.01 | 0.907 | 10.66 | 11.34 | 0.816 |
LSTM | 3.72 | 3.78 | 0.942 | 4.25 | 4.42 | 0.929 | 9.43 | 10.06 | 0.834 |
XGBoost | 3.85 | 3.81 | 0.949 | 4.42 | 4.59 | 0.897 | 9.71 | 10.27 | 0.823 |
BiLSTM | 3.58 | 3.67 | 0.955 | 3.94 | 4.02 | 0.945 | 8.68 | 9.13 | 0.847 |
CNN-LSTM | 3.23 | 3.29 | 0.963 | 3.53 | 3.66 | 0.951 | 7.54 | 8.28 | 0.856 |
CNN-BiLSTM | 2.95 | 2.92 | 0.971 | 3.28 | 3.41 | 0.962 | 6.95 | 7.62 | 0.868 |
CNN-BiLSTM-Att | 2.73 | 2.71 | 0.988 | 3.03 | 3.12 | 0.979 | 6.33 | 6.82 | 0.875 |
