Prediction of Heart Rate Abnormalities Using Data Mining Techniques

Heart rate arrhythmias have become one of the most popular heart issues in human beings at recent years .It has opened a way to many researchers to classify the heart beats issues in the medical field for detecting the arrhythmia signals at early stages which will help in reducing the death rates. The research focuses on classifying the arrhythmias in the ECG signal for the 5 different categories of signals in the MIT-BIH dataset. A lot of research has been done in the classification for arrhythmias signal using various data mining techniques. Logistic Regression and CNN algorithms were implemented to classify the heart beat signals and aims at providing better accuracy in predicting the signals. Both the implemented models have been compared in terms of accuracy, precision, recall and computational time. The evaluation results of the deep leaning model - CNN turned out to be high in accuracy of 99.09% when compared with the machine learning model. This research helps out in classifying the heartbeat rate arrhythmias at early stages for enhanced treatment. Keywords: Arrhythmias, Wavelet Transformation, Logistic Regression, CN