Real Time EEG-based Stress Detection and Meditation Application with K-Nearest Neighbor
P. D. Purnamasari, Alya Fernandya
2019 IEEE R10 Humanitarian Technology Conference (R10-HTC)(47129) November 1, 2019 DOI: 10.1109/r10-htc47129.2019.9042488 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Applied system development and classification study Peer reviewed |
|---|---|
| Topics | Meditation |
| Key points | The authors report that delta, theta, alpha, and beta EEG waves were the most suitable features for stress detection, and that a k-Nearest Neighbor classifier with k = 3 achieved the best performance at 80% accuracy. They propose the resulting application for both detecting stress and supporting meditation-based stress reduction. |
Abstract
The desire to end someone's life commonly triggered by depression or mental health. One way to detect whether someone is depressed is by checking whether the person is constantly stressed for a long time. Electroencephalograph (EEG) is one of physiological signal that can be used for monitoring someone's mental condition, such as stress. The proposed and developed Stress Detection and Meditation Application is an application that can be used for meditating and reducing stress level by using EEG to get brainwave signals. In this developed system, Fast Fourier Transform (FFT) was used as the feature extraction and k-Nearest Neighbor (k-NN) was used to classify the features and detect whether the person is stressed or not. The Delta, theta, alpha, and beta waves as the features were the most suitable feature type for the stress detection and meditation application, while the value of $k = 3$ in k-NN classification provides the best $k$ value with 80% accuracy.