Performance analysis of eye-state charecterization through single electrode EEG device for medical application

Abstract

Electroencephalography (EEG) based human computer interface for upgrading the personal satisfaction is one of the burning research field in restorative and additionally non-medicinal applications. Such innovation can be consolidated to brain science, anesthesiology, gaming, security framework and for continuous patients monitoring. The Neurosky Mind wave headset gadgets are for the most part use to identify and measure electrical action of the client's temple and transmit the gathered information remotely, to a Computer for further preparing. Subsequent to preparing EEG signal, it is classifications into different recurrence groups for highlight extraction. This paper deals with Eye state prediction of dataset using different classifiers viz. Ada Boost, Naïve Bayes and Multilayer perception (MLP) through WEKA. In this paper we are acquire different subjects from the age gap between 18–40 years perform Eye state levels have been examined. We have used different machine learning schemes for Eye blinking and Eye Open but MLP classifier provides highest classification rate i.e. 85% and ROC [Receiver Operating Characteristics] area is 0.855.