Event-Related Potentials Signal Feature Classification Algorithm Based on Genetic Algorithm

To solve the problem of insufficient feature information obtained by the single feature extraction method. The feature extraction was achieved by autoregressive model (AR) and wavelet transform. After the merging of feature sets, the genetic algorithm is used to select the optimal feature set. To test the validity of the proposed method, it is compared with the feature selection method based on SGA and the filter selection method based on Fisher distance. The classification accuracy of AGA is significantly higher than other methods, and the best pattern recognition performance is obtained. SGA-FS and AGA-FS take classification accuracy as the index to evaluate different feature subsets, so it is expected to obtain better classification results.