Middle Phase of Seizure and Seizure-free EEG Signals Classification Using Fractional Pseudo Controller: Linear Forecasting Approach

In this paper, a technique is shown for electroencephalogram (EEG) signal grouping based on the fractional-arrange mathematics. This technique, named as the fractional linear forecasting (FLF) is utilized to display the middle phase of seizure (Ictal Pahse) and seizure EEG signals. It is discovered that the displaying blunder vitality is high considerably for ictal EEG signals contrasted with sans seizure EEG signals. In addition, it is realized that middle phase of seizure (Ictal) EEG signals have higher energy than sans seizure EEG signals. These two parameters are then given as contributions to prepare a support vector machine (SVM). The prepared SVM is then used to group an arrangement of EEG signals into middle phase of seizure (Ictal) and without seizure classifications. It is discovered that the proposed technique gives an order forecasting of 95.33% when the SVM is organized with the Radial Basis Function (RBF) kernel.

Keywords:

EEG Signal , Seizure , Support vector machine , Fractional Forecasting