Optimized SVM based on improved whale algorithm for EEG signal pattern classification
Junqiang Peng, Lijun Liu · 2023
In order to improve the classification accuracy of moving image electroencephalogram ( EEG ) signals, aiming at the problem that the traditional support vector machine (SVM ) parameter value is not accurate, which leads to poor classification model, this study introduces Sobol sequence, reverse learning strategy and piecewise nonlinear time-varying factor. In order to improve the random layout strategy of boundary processing, Levy flight strategy improves the traditional whale algorithm, combines the improved whale algorithm with SVM, and compares the combined SVM classifier with other classifiers for EEG signal classification. The results show that the accuracy of the optimization algorithm is 2 ~ 4 percentage points.