Classifying Single-Trial EEG during Motor Imagery with a Small Training Set

TL;DR

This paper presents methods to classify single-trial EEG during motor imagery with limited training data, achieving high accuracy and addressing overfitting challenges for practical brain-computer interfaces.

Contribution

It introduces a combination of signal processing and machine learning techniques specifically designed for small training sets in EEG classification.

Findings

Achieved test accuracy comparable to training accuracy.

Successfully addressed overfitting in small-sample EEG classification.

Validated methods on BCI Competition III data and own experiments.

Abstract

Before the operation of a motor imagery based brain-computer interface (BCI) adopting machine learning techniques, a cumbersome training procedure is unavoidable. The development of a practical BCI posed the challenge of classifying single-trial EEG with a small training set. In this letter, we addressed this problem by employing a series of signal processing and machine learning approaches to alleviate overfitting and obtained test accuracy similar to training accuracy on the datasets from BCI Competition III and our own experiments.