Lead optimization method for SVM-RFE (support vector machine-recursive feature elimination) based on ensemble learning thought

Abstract

A lead optimization method for a P300 brain computer interface of SVM-RFE (support vector machine-recursive feature elimination) based on ensemble learning thought includes the steps: performing data acquisition through the visual P300-Speller BCI (brain computer interface), and pre-processing the acquired data; determining the sample number of a basic learner in ensemble learning; and performing recursive feature elimination of the support vector machine based on the ensemble learning thought. The method integrates the ensemble learning thought into the SVM-RFE, so that lead optimization can be effectively completed, calculated amount during lead optimization is greatly reduced while the optimization quality is ensured, improvement on online degree of the BCI system and promotion of commercialization of the BCI system are facilitated, and considerable economic and social benefits can be obtained in the man-machine interaction field.

Description

technical field [0001] The invention relates to a P300 brain-computer interface lead optimization method. In particular, it involves integrating the idea of ​​ensemble learning into the Recursive Feature Elimination Base on Support Vector Machine (SVM-RFE), which greatly reduces the calculation of lead optimization while ensuring the quality of optimization. Quantitative SVM-RFE lead optimization method based on the idea of ​​ensemble learning. Background technique [0002] BCI realizes a new way of external information exchange and control by using engineering technology to "turn thought into action". It is a cross-cutting technology involving medicine, neurology, signal detection, signal processing, pattern recognition and other fields. The BCI system can not only help patients with motor dysfunction but normal thinking to communicate with the outside world through the thinking movement of the brain, but also has important potential value in virtual reality, game entertai...

AI Technical Summary

Problems solved by technology

While a large number of leads cover more abundant EEG information, it also brings many problems. For example, a large number of leads brings a huge amount of data, which brings a great burden to the subsequent signal processing. Serious Affected the real-time performance of the system

In addition, the more leads means the greater the possibility of introducing interference, affecting the stability of the system

In practical applications, too many leads will seriously affect the operability of the system, reduce the comfort of the user, and greatly reduce the practicability of the system

On the other hand, due to the existence of individual differences, there is no general lead combination that can ensure that all users can achieve better results, which requires the use of lead screening to personalize lead for different individuals. set up

Images