ECG-Based Prediction of Epileptic Seizures Using Machine Learning Methods

Epilepsy is a type of neurological disorder that is associated with recurrent seizures. This study aimed to present three machine learning methods for predicting epileptic seizures using the features of electrocardiogram (ECG) signals. First, the ECG data set consisting of 13 people were pre-processed. Then, 13 features were extracted. SVM, KNN, and Naive Bayes were used to classify the interictal and preictal sections of each patient's ECG signal. The prediction interval of 15 minutes was selected. The results show that the Naive Bayes is a suitable choice to predict epileptic seizures.