Research Article
Analysis of the Oil Content of Rapeseed Using Artificial Neural Networks Based on Near Infrared Spectral Data
Table 2
Results of different models in analyzing oil content of rapeseed.
| | Model | Trained samples | Tested samples | RMS error | Stopped reason |
| | Linear prediction | 24 | 5 | 0.60 | Autostopped | | GRNN | 24 | 5 | 0.74 | Autostopped | | MLFN 2 nodes | 24 | 5 | 1.04 | Autostopped | | MLFN 3 nodes | 24 | 5 | 1.23 | Autostopped | | MLFN 4 nodes | 24 | 5 | 0.88 | Autostopped | | MLFN 5 nodes | 24 | 5 | 1.29 | Autostopped | | MLFN 6 nodes | 24 | 5 | 1.31 | Autostopped | | MLFN 7 nodes | 24 | 5 | 2.20 | Autostopped | | MLFN 8 nodes | 24 | 5 | 0.59 | Autostopped | | MLFN 9 nodes | 24 | 5 | 3.39 | Autostopped | | MLFN 10 nodes | 24 | 5 | 1.93 | Autostopped | | MLFN 11 nodes | 24 | 5 | 0.83 | Autostopped | | MLFN 12 nodes | 24 | 5 | 1.54 | Autostopped | | MLFN 13 nodes | 24 | 5 | 1.10 | Autostopped | | MLFN 14 nodes | 24 | 5 | 1.57 | Autostopped | | MLFN 15 nodes | 24 | 5 | 3.02 | Autostopped | | MLFN 16 nodes | 24 | 5 | 1.67 | Autostopped | | MLFN 17 nodes | 24 | 5 | 1.07 | Autostopped | | MLFN 18 nodes | 24 | 5 | 1.72 | Autostopped | | MLFN 19 nodes | 24 | 5 | 0.85 | Autostopped | | MLFN 20 nodes | 24 | 5 | 2.79 | Autostopped |
|
|
