Impulsion Assessment and Classification Based on EEG Features

EasyChair Preprint no. 9030, version 2

Abstract

The study describes a system which integrates EEG power spectrum, power spectrum density (PSD), sample entropy and other features in the β frequency band to evaluate, predict and verdict in an emotion classifier. The system proposed an accurate classification method based on EEG spectrum imported into SVM classifier for β wave. In the experiment, the EEG data of subjects in resting state, mild impulsive environment and high impulsive environment were extracted. From the variance analysis results, there was no significant difference in the β power spectral density between the resting state and the mild pulse environment (p = 0.089), but the β power spectral density changed significantly in the mild impulsive environment and the high impulsive environment (p < 0.001). The significant difference in the power spectrum of the β band under the two different states was successfully given, which has guiding significance. After power spectrum transformation, Support Vector Machine (SVM) classifier is used for classification. Our classifier showed an average accuracy of 88.46%.

Keyphrases: Classifier, EEG power spectrum, feature selection, impulsive state, β band

@Booklet{EasyChair:9030,
  author = {Zhanhao Jin and Xin Xu},
  title = {Impulsion Assessment and Classification Based on EEG Features},
  howpublished = {EasyChair Preprint no. 9030},

  year = {EasyChair, 2022}}