Applying Data Mining Classification Techniques and Ensemble Learning to Identify Epileptic Pediatric

Abstract

Epilepsy is the second most common neurological disease after Alzheimer’s. It is a disorder of the brain that causes recurrent seizures. The Effects of epilepsy in children are serious, as it causes a slower rate of development and a failure to improve certain skills. Seizures are the most common symptom of epilepsy. In the medical field, specialists record brain activity using an electroencephalogram (EEG) to observe epileptic seizures. The detection of these seizures is performed by specialists, but the results might not be accurate and depends on the specialist’s experience; therefore, automated detection of epileptic pediatric seizures might be the optimal solution. This research study investigates the detection of epileptic seizures by applying supervised machine learning techniques, the techniques applied on the patients with 7 years old and less from CHB-MIT scalp EEG database of epileptic pediatric signals. In addition, the EEG data were analyzed to clean it of artifacts and extract useful features to classify them as an epileptic or non-epileptic seizure, which was conducted via MATLAB. A decision tree, logistic regression, linear discernment, support vector machine, k-nearest neighbor, naïve Bayes, and ensemble learning methods were applied to the classification process. The results demonstrated the outperformance on existing studies by achieving 100% for all parameters for the Ensemble learning model. The SVM model achieved good performance with 98.3% for sensitivity, 97.7% for specificity, and 98% for accuracy. The results of the LD and LR models reveal the lowest performance: the sensitivity at 66.9%–68.9%, specificity at 73.5%-77.1% and accuracy at 70.2%-73%.