An intelligent model for ECG Classification System based on frequency domain with Least Square Support Vector Machine (LS-SVM)

Abstract

The development of authentication and identity mechanisms has become a vital requirement to secure device data integrity, although passwords provide sufficient control and authentication, they have shown major flaws in speed and security, Biometrics is becoming the primary authentication method. Electrocardiogram (ECG) signals are created as a consequence. Due to their unique character, which makes them difficult to falsify and ubiquitous, ECG signals have attracted much attention in most authentication systems. We introduce a novel ECG validation model that combines frequency domain-based characteristics with the least-squares technique in this work (LS-SVM). Two types of frequency field characteristics were examined in order to find the best combination of high precision and speed. ECG Signal Characteristics and Frequency Domain Characteristics ECG signals are used to determine the optimal triple-band filter bank. To access the most important characteristics, we removed the extraneous features and extracted others.