Table 2
Description of the best experiments for the 18 bands (Exp. n. 37) and the 14 bands datasets (Exp. n. 26).
| Exp. n | Missing features | Feat. | Hid. | Step | Res. | Dec. | MxIt | CV | Scatter | Outliers% | Bias |
|
|
|||||||||||
| 37 | m5.8, K, HK, m8 | 14 | 29 | 0.0001 | 30 | 0.1 | 3000 | 10 | 0.057 | 22.61% | –0.0077 |
| 26 | m5.8, K, m3.6, m4.5, HK, m8 | 12 | 25 | 0.0001 | 30 | 0.1 | 3000 | 10 | 0.062 | 17.39% | 0.0078 |
Notes. Column 1: sequential experiment identification code; Col. 2: features not used in the experiment; Cols. 3–4: number of input (features) and hidden neurons; Cols. 5–9: parameters of the MLPQNA used during the experiment; Col. 10: scatter error evaluated as described in the text; Col. 11: fraction of outliers; Col. 12: bias.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.