Outlier Detection System for Cardiovascular disease using Time-Series Data: A Comparative Analysis

Traditional approaches to disease diagnosis and prognosis are cumbersome and prone to error. Cardiovascular illness is extremely serious and needs to be handled by qualified healthcare professionals. Any small error causes huge problems. Using the conventional methods, it is extremely difficult to diagnose and forecast cardiovascular disease in its early stages. Artificial intelligence has the power to solve a wide range of issues in a variety of fields, with healthcare being one of the most promising. The most reliable approach for assessing the performance and health of the cardiovascular system, as well as for detecting disorders related to it, is electrocardiogram (ECG) signals. The arrhythmias that are represented by ECG patterns are irregular heartbeats. A growing amount of the ECG arrhythmia signal needs to be automated for classification and identification in order to produce quicker and more accurate findings. To evaluate the efficacy and accuracy of the outlier detection method over benchmark datasets towards the model, the Arrhythmia dataset and the PTB abnormal normal dataset were used in this work.