PERFORMANCE ANALYSIS OF DIMENSIONALITY REDUCTION TECHNIQUES IN EEG SIGNAL CLASSIFICATION

Dimensionality Reduction is paying more consideration now a day because of rising need to effectively manage massive volumes of data. The applications of Dimensionality Reduction cover many fields like Medical, Geographical, E-Commerce, simulation and many more. Dimensionality Reduction (DR) plays an important part in selecting essential features. So that it reduces the dimensions of the dataset. The Dimensionality Reduction techniques were used to select the most relevant features of face recognition, Speaker identification, image annotation and many more. A proposed method for classification of the EEG signals is explained in this paper. It is based on DWT, the techniques for dimensionality reduction and SVM classification. Using DWT, the EEG signals are divided into the frequency sub-bands. Then a number of statistical features were extracted from the sub-bands to represent the distribution of the wavelet coefficients. Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Independent Component Analysis (ICA) are used to reduce data dimensions. After reducing, these features were finally given as an input to a support vector machine (SVM) with two discrete outputs. They are epileptic and normal seizure. The performance of dimensionality reduction process is presented and compared with the proposed method to show the excellence of classification process in EEG signal.