Skip to content

Code to perform experiments predicting the movement direction of a subject based on EEG data

Notifications You must be signed in to change notification settings

tylerfeldman321/Brain-Computer-Interface-Movement-Decoding

Repository files navigation

Brain-Computer-Interface-Movement-Decoding

This repo contains code for a machine learning project to decode if a person would like to select a left vs. right option based on the electroencephalogram (EEG) signals from the person's brain. For detailed information about the modeling approach, results, etc., see the included pdf file: Brain-Computer-Interface-Movement-Decoding.pdf

Modeling Approach

We used a support vector machine to classify left vs. right intended movement based on the EEG signals from the subject. An SVM was used to handle the high-dimensionality of dataset. Many improvements were made upon a basic SVM, which are discussed in the attached pdf file.

Dataset

To evaluate the modeling approach, we used two different datasets. The first dataset, which we called "overt", contains data from subjects who were actually moving thei left or right hands to select an option. The second dataset, which we called "imagined", contains data where the subjects are imagining moving their left or right hand, but not actually moving them.

About

Code to perform experiments predicting the movement direction of a subject based on EEG data

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages