Confidence-based Early Classification for Motor Imagery Brain-Computer Interfaces

Hassan, Mohamed (2024) Confidence-based Early Classification for Motor Imagery Brain-Computer Interfaces. Bachelor's Thesis, Artificial Intelligence.

Preview	Text bAI_2024_HassanM.pdf Download (1MB) | Preview
	Text toestemming.pdf Restricted to Registered users only Download (88kB)

Abstract

Brain-Computer Interface (BCI) research has seen advancements in recent years with regard to early classification and confidence. This project explores a confidence-based early classification method to improve the earliness and performance of Motor Imagery (MI) BCI. A dynamic early classification model was implemented using a stopping criterion based on prediction confidence to determine the optimal timing for classification decisions. This was compared to a static classification model by providing the optimal early stopping times found by the dynamic model. The models were tested on the BCI Competition IV dataset 2a (Brunner et al., 2008), using Linear Discriminant Analysis (LDA) and Support Vector Machine (SVM) as classifiers. The models were evaluated on Accuracy, Cohen’s Kappa, and Information Transfer Rate (ITR). Results showed increased or maintained performance for the confidence-based early classification models as compared to the static model.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Jong, I.P. de
Degree programme:	Artificial Intelligence
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	30 Jul 2024 13:57
Last Modified:	31 Jul 2024 09:56
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/33772

Actions (login required)

View Item