Table 12.
LGG dataset: SVM and random forest are based on stacked views. Deep IDA + SVM means selecting features from Deep IDA and training an SVM classifier on these features. iDeepViewLearn with selected top 50 features obtains a classification error based on a shared low-dimensional representation trained on data with the selected top 50 features. Similar for iDeepViewLearn with selected top 100 features
| Method | AverageError (%) |
|---|---|
| SVM on stacked data | 30.00 |
| Random Forest on stacked data | 26.00 |
| iDeepViewLearn with selected top 50 features | 28.00 |
| iDeepViewLearn with selected top 100 features | 26.00 |
| SIDA | 29.00 |
| Deep GCCA + SVM | 29.00 |
| Deep IDA | 28.00 |
| Deep IDA + SVM | 26.00 |