Feature extraction and performance analysis of EEG signal using S-transform

Abstract

Feature can be described as a functional component observed from a data set. The extracted features give the information related to a signal, thus it requires to calculate cost of information processing and complexity of analyzing a huge data set. This paper presents a feature extraction method using S transform. Five data sets are taken and feature extraction has been performed by implementing two methods: first by applying S-transform and other without S-transform. The performance of the neural model is evaluated on the basis of training performance and classification accuracies and the results confirmed that the proposed scheme has potential in classifying the EEG signals.