A novel combination of code converters and sparse representation classifiers for an efficient epilepsy risk level classification

The aim of this paper is to give a performance analysis by considering the advantage of Code Converters as a feature extraction technique and Sparse Representation Classifier (SRC) as a post classifier for the classification of the epilepsy risk levels obtained from Electroencephalography (EEG) signals. A group of related or similar disorders which is generally characterized by the occurrence of frequency and recurrent seizures is termed as epilepsy. There are different types of epilepsy and seizures. To control the seizures, epilepsy drugs are usually prescribed and if medication is ineffective, then surgery can be done to control or prevent the occurrence of the seizures. This paper aims at the classification of epilepsy risk levels using Sparse Representation Classifier. The Performance Index (PI) and Quality Values (QV) are the two parameters that are used to assess the performance of the code converter and the sparse representation classifiers.