Research Article
A Time-Varying Seasonal Autoregressive Model-Based Prediction of Respiratory Motion for Tumor following Radiotherapy
Table 1
Tested prediction methods and settings.
| Method | Setting |
| ZOH | None | — |
| | Samples back length | 300 samples | SSA [19] | Dimension number of covariance matrix | 250 samples | | Number of largest eigenvalues | 18 |
| | Sampling interval | 15 samples | KDE [18] | Dimension number of covariate | 3 | | Length of moving window | 300 samples | | Representing value of distribution | Mean |
| Adaptive SAR [17]
| Order of seasonal AR | | | Coefficients of SAR | , |
| | Order of TVSAR | | TVSAR | Coefficients of TVSAR | , | (proposed method) | Reference interval estimation | (a) Correlation analysis-based approach | | | (b) Adjustment with adjustment range |
|
|