Table 3 Performance of the proposed CFNN using varying train/test splits.

From: Detecting emotions through EEG signals based on modified convolutional fuzzy neural network

Train/Test

Results

Accuracy*

Precision

F1-score

Recall

90:10

96.94% (96.86–97.01)

95.93%

96.43%

96.93%

80:20

96.85% (96.78–96.92)

96.25%

96.49%

96.73%

75:25

97.24% (97.18–97.30)

96.47%

96.79%

97.12%

K = 5

97.79% (97.71–97.87)

97.70%

97.67%

96.98%

  1. *Values in parentheses indicate a 95% confidence interval; bold values indicate the best values, while italic values indicate the second highest.
Back to article page