Pattern Classification of EEG Brainwave Signals Under the Influence of High Frequency RF Radiation

The effects of mobile phone usage on human health are now becoming the subject of recent interest and study. Analysis and observations of the electroencephalogram (EEG) signals can provide valuable insight and thus improves the understanding of the radiofrequency (RF) radiation influence towards human brain. This paper evaluates the selected classifiers for the classification of brainwave datasets due to the effects of RF radiation. The classification techniques considered here are Discriminant Function Analysis (DFA), Logistic Regression (LR), k-Nearest Neighbor (kNN) and neural network Back Propagation (BP). Beta, alpha, theta and delta brainwaves were used as inputs to the classification system with three discrete outputs: Left Exposure (LE), Right Exposure (RE) and Sham Exposure (SE) group. These classifiers are evaluated based on the classification accuracy and the number of samples correctly classified. The BP based classifier outperformed the other classifiers with 100% classification accuracy. Keywords— EEG, brainwave, asymmetry, radiation,