Table 3.
State of the art comparison.
| Validation scheme | Database: Emotional model | Accuracy (%) | |
|---|---|---|---|
| Khosrowabadi et al. (2014) | Cross-validation | DEAP: valence/arousal | 70.83/71.43 |
| Zheng and Lu (2015) | IT | SEED: [pos., neg., neu.] | 86.65 |
| Zheng et al. (2017) | LOO and IS | SEED: [pos., neg., neu.] | 60.93 and 79.28 |
| Tripathi et al. (2017) | LOO | DEAP: valence/arousal | 66.79/57.58 |
| Song et al. (2018) | IT | SEED: [pos., neg., neu.] | 90.4 |
| Liu et al. (2017) | Not specified | Own produced: [neu., non-neu.]/[pos., neg.]/[joy, amusement, tenderness]/ [sad, angry, fear, disgust] | 92.26/86.63/86.43/65.09 |
| Proposed method | IT/IS/LOO | SEED: [pos., neg., neu.] | 82.27/76.36/77.59 |