Key result
Support Vector Machine with RBF kernel achieved an overall accuracy of 91% in classifying EEG signals of normal, poor dyslexic, and capable dyslexic children when kernel scale and box constraint were set to one.
Cross-Sectional (n=33)
An SVM classifier using an RBF kernel with optimized parameters can accurately classify EEG signals of normal, poor dyslexic, and capable dyslexic children with 91% accuracy.
Dyslexia is referred as learning disability that causes learner having difficulties in decoding, reading and writing words. This disability associates with learning processing region in the human brain. Activities in this region can be examined using electroencephalogram (EEG) which record electrical activity during learning process. This study looks into performance of Support Vector Machine (SVM) using RBF kernel in classifying EEG signal of Normal, Poor and Capable Dyslexic children during writing words and non-words. Discrete Wavelet Transform (DWT) with Daubechies order 2 was employed to extract the power of beta and theta waves of EEG signal. Beta and Theta/Beta ratio form the input features for classifier. Multiclass one versus one SVM was used in the classification where RBF kernel parameters and box constraint values were varied with the factor of 10 to analyze performance of the classifier. It was found that the best performance of SVM with 91% overall accuracy was obtained when both kernel scale and box constraint are set to one.
No takes yet. Share an insight, caveat, or question.
Zainuddin et al. (2018) conducted a cross-sectional in Dyslexia (n=33). Support Vector Machine (SVM) with RBF kernel was evaluated on Overall classification accuracy. Support Vector Machine with RBF kernel achieved an overall accuracy of 91% in classifying EEG signals of normal, poor dyslexic, and capable dyslexic children when kernel scale and box constraint were set to one.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: