Seminar: Linear Discriminant Analysis of Brain Neural Signals

Bukunola Ladele
Honours Thesis Project
Supervisor: Dr. Edward Brown

Linear Discriminant Analysis of Brain Neural Signals

Thursday, April 2, 2015, 1:30 p.m., Room EN-2022
Department of Computer Science


Abstract

Electroencephalogram (EEG) headsets are becoming cost-effective for individual applications and personal use. Some of these applications include usage as an input device for physically challenged users, but there are many other possibilities. However, the ability of consumer grade EEG sensors to reliably discern different mental activities is still a matter of empirical investigation, and the algorithms used by hardware manufacturers remain proprietary and undisclosed. This presentation aims to provide an introduction to how data from EEG headsets can be classified for use as input devices with a focus on Linear Discriminant Analysis (LDA), a simple yet powerful machine learning algorithm.