EEG-ECG Signals Classification for Arrhythmia Detection using Decision Trees

The study analysis the electrocardiogram (ECG) and electroencephalogram (EEG) signals classification problem by using decision trees. The analysis and classification of heartbeats and brain traces associated with different types of Arrhythmia and Seizure is an active research topic in recent years. In this paper, we discuss different classification techniques for the analysis and classification of ECG and EEG signals. The training and testing is performed on the MIT-BIH arrhythmia and ECG database. For classification, the random forest tree and naïve Bayes algorithms are used. The results provide an improved performance over the current standards used for classification. The accuracy of the designed system is 97.45% which outperforms the various instance learning and supervised machine learning methods.