Description

Electroencephalography signal is the recording of electrical activity of brain, provides valuable information of the brain function and neurological disorder. Classification of the EEG signals helps in the identification of emotions in persons or abnormalities in persons. Classification includes training and testing steps. In training step the features were extracted for some set of signals. The classification process is employed using machine learning techniques like j48 and KNN. Machine learning techniques were capable of the identification of the classes in the input data based on the feature values extracted from the signals. For extraction of features Empirical Wavelet Transformation (EWT) process is employed. The input EEG signals were obtained from dataset. The features were extracted from the signals based on Empirical Wavelet Transform (EWT). EWT decompose EEG signal into a finite sum of Intrinsic Mode Functions (IMF) which is an AM-FM signal. The statistical values were extracted from the decomposed signals resulting in the EWT features. A method that classifies the input EEG signals based on the wavelet features is proposed. The features were extracted from EEG signal based on the Empirical wavelet transform. The EEG signals were decomposed using EEG signals. In EWT the levels of decomposition is based on input. Extracted features from signals is more effective compared to the features extracted from DWT From the decomposed signals features were extracted based on the statistical features like mean and standard deviation. The extracted features were classified using classifiers like j 48 and KNN classifier inorder classify the signals. The performance of the process is measured based on performance metrics like Accuracy, Sensitivity and Specificity of the classifier. Our final goal of the study is the automatic detection of the epileptic disorders in the EEG in order to support the diagnosis and care of the epileptic syndromes and related seizure disorders.