Skip to main content

Table 5 The one-step and multistep forecasting accuracy of the ARIMA and XGBoost models

From: Time series analysis of hemorrhagic fever with renal syndrome in mainland China by using an XGBoost forecasting model

Model
Strategy
Index
ARIMA XGBoost
One-step Multistep One-step Multistep
Training set Test set Training set Test set Training set Test set Training set Test set
ME − 7.149 − 61.448 − 7.149 − 259.878 8.111 33.622 8.111 97.931
RMSE 181.977 249.276 181.977 302.781 166.311 178.547 166.311 223.187
MAE 108.160 185.367 108.160 259.878 113.219 132.055 113.219 173.403
MPE − 0.937 − 6.575 − 0.937 − 30.121 − 2.403 2.383 − 2.403 6.348
MAPE 10.293 18.561 10.293 30.121 11.596 12.353 11.596 15.615
MASE 0.442 0.757 0.442 1.062 0.462 0.526 0.462 0.691
ACF1 0.016 − 0.169 0.016 − 0.159 0.424 − 0.232 0.424 − 0.047
Theil’s U NA 0.375 NA 0.441 NA 0.273 NA 0.398