Innovative Approach for Early Heart Disease Prediction: Gaussian Mixture Model and K-Nearest Neighbour

An Electrocardiogram (ECG) contributes significantly to early diagnosis and classification of heart diseases, arrhythmia which means irregular heart rate. Regrettably, the process became difficult due to asymmetric and minimal amplitude from recordings of ECG and cannot do the extraction from multiple signals. By using machine learning technology, it can classify and extract multiple samples, but cannot extract the data features from complex data structures. To overcome the limitations, previously introduced metaheuristic optimization (MHO) to improve the classification and extraction. Proposed a method which supports and K-Nearest Neighbour (KNN) Gaussian Mixture Model (GMM) algorithms for determine the abnormalities, predicting cardiac events, and improve diagnostic accuracy. As a result, this can help us to get results more quickly and accurately with sensitivity 99.7% and accuracy 99.2%, for helping healthcare professionals in diagnosing arrythmias, myocardial infarction and any other heart conditions.