Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/74813
Title: Convolutional neural network based two-stage epileptic EEG classification system
Authors: Gan, Sie Huai
Keywords: DRNTU::Engineering
Issue Date: 2018
Abstract: Epilepsy is a central nervous system disorder and the epileptic patients will exhibit ‘spikes’ characteristics in their brain wave. The objective of this project is to discover a better setup parameters of the Convolutional Neural Network (CNN) and techniques in improving the classification result of the automated epileptic patient classification system. Different architecture has been experimented on the system and the system was evaluated based on several metrics. Besides that, various of techniques such as background data swapping, early stopping, training with different background to spikes ratio and so on were also attempted in this project and the effect on the improvement of the system has been investigated. In conclusion, a better setup parameter and techniques has been discovered and achieve improvement to the classification accuracy of the system. Further improvement may include training with more patient data and use different techniques such as batch normalization and pseudo-labelling to improve the accuracy of the system.
URI: http://hdl.handle.net/10356/74813
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 
Final Year Report (Dr NTU version).pdf
  Restricted Access
2.01 MBAdobe PDFView/Open

Page view(s) 50

85
checked on Oct 24, 2020

Download(s) 50

7
checked on Oct 24, 2020

Google ScholarTM

Check

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