Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/72041
Title: Machine learning algorithm for electroencephalography (EEG) based brain signal analysis
Authors: Teo, Jeffrey Eng Hock
Keywords: DRNTU::Engineering::Electrical and electronic engineering
Issue Date: 2017
Abstract: Emotions are one of the many ways humans communicate with one another. One of the ways to record these emotions is by collecting brain signals via electroencephalography (EEG). There are many applications that this can be used for, be it in the medical field or for artificial intelligence purposes. The main objective of this study is to develop a machine learning algorithm that can automatically predict the emotional state (Happy or Sad) of a human, based on the EEG signals. This is done in 3 stages. The first stage involves the extraction of features in the alpha and beta band, from data that have been collected. The feature extracted is Discrete Wavelet Transform approximate coefficient at level 1. The next stage is to select the features using the Fisher’s ratio. The final stage is to use various classification methods to classify the data and test the accuracy of the models. The 3 classifiers being evaluated are Linear Discriminant Analysis (LDA), Linear Support Vector Machine (SVM) and k-Nearest Neighbour (KNN). A 6-fold cross validation was used. After evaluation of the 3 classification methods, it was found that LDA works best with an accuracy of 92.9% in the alpha band. For beta band, KNN gives the best prediction accuracy of 92.9%. When analysing both alpha and beta bands, KNN was found to be the best classifier to predict the emotional state (Happy or Sad).
URI: http://hdl.handle.net/10356/72041
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
FYP_Report_Teo_Eng_Hock_Jeffrey.pdf
  Restricted Access
Final Year Project Final Report2.23 MBAdobe PDFView/Open

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.