ECG Classification System based on time Domain Features with Least Square Support Vector Machine (LS-SVM

The development of authentication and identification procedures has become a critical requirement toensure the integrity of device data. Although passwords provide enough control and authentication, theyhave been shown to have significant speed and security flaws, making biometrics the preferredauthentication technique. Consequently, most authentication systems have given much consideration toan electrocardiogram (ECG) signal. However, ECG signals are distinct, making them difficult to imitateyet relatively frequent. We offer a novel ECG validation model that combines multi-domaincharacteristics with an LSSVM in this work. ECG signals examine two characteristics to determine eachindividual's ideal mix of characteristics. First, the best 3-band filter bank is used to extract time-domainproperties from ECG data. The retrieved characteristics are examined for the most relevant ones, and theunimportant ones are removed. The Least Square Support Vector Machine(LS-SVM) classifier collectsspecified characteristics data. Our ECG authentication technology outperformed the competition,according to the findings. The suggested model was applied to categorize EKG signals, and it achieved anaverage accuracy of 92 %and 91 %when time features were used. The suggested model is evaluatedbased on its ability to interact with open data.