EEG Data Analysis for Stress Detection
Lokesh Malviya, Sandip Mal, Praveen Lalwani
Abstract
An electroencephalograph (EEG) tracks and records brain wave sabot. A little size of Metal discs called electrodes. It is connected with wires and used to collect electrical impulses in the brain. Afterward, collected signals forwarded and store using a computer application. Analysis of Stress Levels in a human while performing different tasks is a challenging problem that can be utilized in healthcare systems and applications. In the proposed system, a new technique for band power selection using Fast Fourier Transform(FFT) from EEG channels and using various type of machine learning classification algorithm like Random Forest(RF) Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Linear Regression(LR) validated using accuracy. In the performance analysis, an accuracy of 78.6% is achieved using RF which is higher than other tested machine learning models.