B2 shows slight improvement and is almost insignificant, indicating that the delta 1 features are weak.
← all excerpts
EEG-based harmful brain activity classification using deep learning and feature fusion.
2
—
—
The sentences
Table 8 shows that B1 (MFCC + 1D-CNN) is highly significant, with a p value of 0.005, which indicates that MFCCs have strong feature values.