tsta

فارسی

Home / سی و دومین کنفرانس ملی و دهمین کنفرانس بین المللی مهندسی زیست پزشکی ایران

Emotion Recognition from EEG signal using GA-FLANN with Whale Optimization Algorithm

Authors :

Mohammadamir Razmi¹ Pouya Faridfar² Seyed Amirreza Navali Hosseini alavi³

1- Department of Aerospace Engineering, Science and Research branch, Islamic Azad University, Tehran, Iran 2- Department of computer Engineering Mashhad branch, Islamic Azad University Mashhad, Iran 3- Department of computer Engineering Mashhad branch, Islamic Azad University Mashhad, Iran

Keywords :

EEG signal،Emotion recognition،GA-FLANN Classification،Whale Optimization Algorithm (WOA)

Abstract :

Non-verbal cues—such as intentions and emotions—are central to human communication. Electroencephalogram (EEG) signals, which record the brain’s electrical activity, offer a direct and culture-independent basis for emotion recognition. We propose a compact and efficient pipeline that uses the Whale Optimization Algorithm (WOA) as a wrapper-based feature selector and a Genetic-Algorithm-trained Functional Link Artificial Neural Network (GA-FLANN) as the classifier. EEG signals are band-pass filtered (4–40 Hz), segmented, and transformed into spectral and connectivity descriptors; WOA searches for an optimal sparse subset of discriminative features that maximizes classification performance, while GA-FLANN performs the final emotion classification. Experiments were conducted on the SJTU Emotion EEG Dataset (SEED) comprising 15 subjects, each with 64 EEG channels sampled at 1000 Hz, and involving three emotion classes (positive, neutral, negative), with comparisons against radial basis function (RBF) and Improved Self-Organizing FLANN (ISO-FLANN) baselines. On SEED, the proposed WOA–GA-FLANN achieves 98.73% accuracy and 98.13% macro-F1, improving over radial basis function (RBF) and Improved Self-Organizing FLANN (ISO-FLANN) by approximately +4.7% and +3.9%, respectively.