Klasifikacija vidnih evociranih potencijala stabilnog stanja u sustavu sučelja mozga i računala

undergraduate thesis

• Title (english): Classification of Steady State Visual Evoked Potentials in Brain Computer Interfaces
• Author: Lucija Planinić
• Mentor: Mario Cifrek (mentor)
• Committee member: Mario Cifrek (predsjednik povjerenstva)
• Committee member: Željka Lučev Vasić (član povjerenstva)
• Committee member: Krešimir Friganović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2018-07-13, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Sučelje mozga i računala je izravna komunikacijska veza između mozga i vanjskog uređaja. Nju ostvarujemo prikupljanjem i obradom signala dobivenih snimanjem električne aktivnosti mozga elektroencefalografom. U ovom radu klasificirani su vizualni evocirani potencijali stabilnog stanja koji se u mozgu pojavljuju pri svjetlosnom podražaju koji treperi određenom frekvencijom. Klasifikacija je česti problem u strojnom učenju te postoje mnoge metode koje se u tu svrhu koriste. Uspoređene su 3 metode koje se koriste za klasifikaciju: linearna diskriminantna analiza, metoda potpornih vektora i neuronska mreža. Od te tri metode najveća točnost predviđanja postignuta je metodom potpornih vektora. Neuronska mreža predviđa sa zanemarivo lošijom točnošću od metode potpornih vektora (razlika od otprilike 1%) a najmanju točnost predviđanja ima linearna diskriminantna analiza.

Abstract (english)

A brain-computer interface is defined as direct communication between a brain and an external device. That is achieved through collecting and processing signals obtained by recording the electrical activity of the brain with an electroencephalograph. In this paper, steady state visual evoked potentials, which appear in the brain when a flickering visual stimulus is presented, were classified. Classification is a common machine learning problem, so there are many methods used for that purpose. Three methods used for classification were compared: linear discriminant analysis, support vector machine and artificial neural network. Between those three methods, the best results were achieved with the support vector machine. The neural network performed with an insignificantly smaller accuracy than the support vector machine (the difference was around 1%), while the linear discriminant analysis algorithm performed with the lowest accuracy.

• Keywords: evocirani potencijali stabilnog stanja; klasifikacija; sučelje mozga i računala
• Keywords (english): steady-state visual evoked potentials; classification; brain-computer interface
• Language: croatian
• URN:NBN: urn:nbn:hr:168:726436
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-03-14 00:01:22