Open AccessElectroencephalography based Emotion Recognition using Fisher’s Linear Discriminant Analysis on Support Vector Machine
Author(s) -
Intan Nurma Yulita,
Dessy Novita,
Asep Sholahuddin,
Emilliano
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1577/1/012004
Subject(s) - support vector machine , linear discriminant analysis , pattern recognition (psychology) , speech recognition , electroencephalography , computer science , artificial intelligence , emotion classification , perceptron , multilayer perceptron , identification (biology) , facial expression , artificial neural network , psychology , neuroscience , botany , biology
Emotions as intense feelings for reactions to something affect someone in interacting with others such as in determining choices, actions, and perceptions. The emotional state of an individual can be seen clearly through facial expression and tone of speech. Apart from facial features or voice features, identification of emotions can also be done through brain waves. This study used an electroencephalogram signal as an input to recognize types of emotions. The electroencephalogram signal was chosen because it can record the true emotions of individuals. The recognition of emotions based on Support Vector Machine (SVM). To improve the performance, this method was combined with Fisher’s Linear Discriminant Analysis (FLDA). The experiments showed the SVM performance increased above 30%. As a comparison, this research also implemented Multi-Layer Perceptron (MLP). The results showed that the performances of SVM and FLDA-SVM were higher than MLP or FLDA-MLP. It showed that FLDA-SVM was the best method of this research in recognizing emotions.