Article Per Year (5 Year)
p-Index From 2019 - 2024
0
P-Index
This Author published in this journals
All Journal Jurnal Nasional Teknik Elektro dan Teknologi Informasi
Evi Septiana Pane
Institut Teknologi Sepuluh Nopember
Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Energy Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Peningkatan Akurasi Pengenalan Emosi pada Sinyal Electroencephalograpy Menggunakan Multiclass Fisher Evi Septiana Pane; Adhi Dharma Wibawa; Mauridhi Hery Purnomo
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 7 No 4: November 2018
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1624.293 KB)

Abstract

EEG signals have a significant correlation to emotions when compared to other external appearances such as face and voice. Due to the low accuracy of emotional recognition through EEG signals, this study proposes a dimensional reduction method for EEG data to address that problem using Multiclass Fisher Discriminant Analysis (MC-FDA). In this study, the experiment was applied on public EEG dataset with three classes of emotions, namely positive, negative, and neutral. Differential entropy features were extracted from the decomposed EEG signals in five frequency band of the delta, theta, alpha, beta, and gamma. The accuracy of emotion recognition was measured using two prevalent classifiers on EEG identification, such as LDA and SVM. To demonstrate the superiority of the MC-FDA method, the PCA dimension reduction method was applied as a comparison. Classification accuracy results from all experiment scenario showed the advantages of the MC-FDA compared to the PCA.The best emotion classification accuracy was obtained from trials on all data from twelve electrodes using the MC-FDA and LDA methods, namely 93.3%. These results show a mean increase in accuracy of 3.5 points from the original feature vector dataset.
Co-Authors Adhi Dharma Wibawa Mauridhi Hery Purnomo