Source Data Selection for Brain-Computer Interfaces based on Simple Features

TL;DR

This paper introduces a transfer learning method for brain-computer interfaces that uses simple covariance-based features to select source data, enhancing performance for new users.

Contribution

It presents the Transfer Performance Predictor method that effectively selects source data using simple features, improving transfer learning in brain-computer interfaces.

Findings

The method outperforms existing source data selection techniques.

Simple covariance-based features are effective for transfer learning.

Improved BCI performance for new users through better source data selection.

Abstract

This paper demonstrates that simple features available during the calibration of a brain-computer interface can be utilized for source data selection to improve the performance of the brain-computer interface for a new target user through transfer learning. To support this, a public motor imagery dataset is used for analysis, and a method called the Transfer Performance Predictor method is presented. The simple features are based on the covariance matrices of the data and the Riemannian distance between them. The Transfer Performance Predictor method outperforms other source data selection methods as it selects source data that gives a better transfer learning performance for the target users.