Table 5 Comparison of classification accuracy across different strategies and models (units: %).
From: Video-dominant emotion recognition for portable EEG-based devices
Models | Classification strategies | ||
|---|---|---|---|
Dynamic hierarchical label calibration | Multi-dimensional energy ratio analysis | Saliency-guided feature selection | |
SVM | 34.65 ± 3.5 | 42.75 ± 3.8 | 43.98 ± 5.5 |
RF | 32.48 ± 7.4 | 40.46 ± 4.2 | 45.82 ± 4.4 |
DNN | 33.08 ± 1.4 | 36.64 ± 4.1 | 41.23 ± 5.9 |
