P300 Analysis using Artificial Neural Network
Electroencephalogram (EEG) is the measurement of electrical activity of the neurons in the brain from the scalp.
This study evaluates the relative performance of two established feature extraction techniques on data collected using the
P300 Speller paradigm, originally described by Farwell and Donchin . We have used the following two methods:
Wavelet Transform (WT) and Principal Component Analysis (PCA) in our research. In this work, WT and PCA are used as a
preprocessing method and neural network is used for classification. With the aim to improve the distinct features extracted
by wavelet transformation in P300 detection, we researched the P300 frequency domain of Event Related Potentials (ERP)
and instigate the mother wavelet selection towards the divisibility of extracted features. PCA has been implemented on P300
for feature reduction for classification.
Index Terms- Brain Computer Interface (BCI), Electroencephalogram (EEG), P300 Speller, Event Related Potential (ERP),
Discrete Wavelet Transform (DWT), Principal Component Analysis (PCA), Neural Network.