Skip to main content

Advertisement

Table 1 Results for GA-based feature selection.

From: Single-trial classification of NIRS signals during emotional induction tasks: towards a corporeal machine interface

Participant No. Common features Selected Across Performance Parameter Sets1 Optimal Parameter Set
   Symbol2 Feature Pair Classification Accuracy3
1 Mean, Skewness LDA-L-20- MeanHbOL 1 MeanHbOL 4 75.00 ± 10.83%
2 Mean, Skewness LDA-L-20+ MeanHbOL 3 MeanHbOL 4 89.67 ± 7.82%
3 Mean, Skewness LDA-L-20+ MeanHbOL 1 MeanHbOL 4 96.67 ± 5.32%
4 Kurtosis, Skewness LDA-L-15- KurtosisHbOL 4 SkewnessHbOL 3 75.33 ± 12.59%
5 Kurtosis, Skewness LDA-L-15- KurtosisHbOL 3 SkewnessHbL 2 88.00 ± 7.93%
6 Kurtosis, Skewness SVM-L-20- SkewnessHbOL 1 SkewnessHbOL 2 75.83 ± 10.55%
7 Mean SVM-L-20+ MeanHbL 4 VarianceHbL 2 94.67 ± 5.77%
8 Mean, Skewness, Ea 6 LDA-R-20+ MeanHbR 3 ZCHbOR 3 89.00 ± 8.82%
9 Mean, Skewness LDA-R-15+ EaHbR 3 SkewnessHbR 3 83.83 ± 9.88%
10 Mean, Skewness, E a LDA-R-20+ Ed6HbOR 3 MeanHbOR 3 78.00 ± 9.78%
  1. 1Found in ≥25% feature pairs across performance parameter sets
  2. 2Symbol defining classification scheme consists of 4 parts: Classifier (LDA/SVM) - Recording Side (L/R) - Analysis Time Interval (15/20) - Stimulus Valence (+/-)
  3. 310 randomized trials, 5-fold cross-validation