An EEG emotion recognition method based on a hierarchical mechanism to build a multi-classifier fusion model

Abstract

The invention relates to an EEG emotion recognition method based on a layered mechanism to build a multi-classifier fusion model, which collects multi-conductor emotion EEG data, and analyzes and processes it, including EEG preprocessing, feature extraction, and weight measurement-based Channel selection to construct emotional EEG feature matrix. The emotional EEG feature matrix is ​​divided into channels according to the electrode position, and the optimal feature selection integration is performed for each channel to construct multiple single emotion classification models. Taking the difference and accuracy obtained by each classification model for the same emotion recognition problem as the evaluation criterion, the optimal single emotion classification model for each channel is selected to obtain the classifier set to be fused. The classification error of each optimal single emotion classification model is used as the weight, and the emotion recognition fusion model is constructed based on the weighted voting method. The invention solves the problem that it is difficult to obtain a higher emotion recognition rate in the EEG sample space by using the fusion of multiple classifiers.

Description

technical field [0001] The present invention relates to the field of emotion computing, and relates to an emotion recognition method based on EEG, in particular to an EEG emotion recognition method based on channel layering mechanism and feature selection integration to build multi-classifier fusion. Background technique [0002] Emotion is an advanced function of the human brain. It is a psychological and physiological state that accompanies the process of cognition and consciousness. It integrates people's feelings, thoughts and behaviors, and plays a very important role in the communication between people. In recent years, with the rapid development of ubiquitous technology and computer technology, emotion recognition, as a key issue of affective computing, has become an important interdisciplinary research topic in the fields of computer science, cognitive science and artificial intelligence, and has received more and more attention. more and more attention and applicati...

AI Technical Summary

Problems solved by technology

[0006] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to propose an EEG emotion recognition method based on a layered mechanism to construct a multi-classifier fusion model, aiming at solving the problem of emotional EEG data with unbalanced, nonlinear and non-stationary categories. When performing classification, the accuracy of existing emotion recognition methods needs to be improved

Images