Performance of various learners on Human Identification based Moving on ECG signal

There is strong evidence that heart’s electrical activity embeds highly distinct characteristics, suitable for applications such as the identification of human subjects. In other words, they contain satisfactory discriminative information to let the identification of individuals from a large population. Therefore, this paper presents a robust identification system using 20 healthy subjects from Physikalisch-Technische Bundesanstalt (PTB) database, 25 subjects from MIT-BIH arrhythmia database and 15subjects from The MIT-BIH Normal Sinus Rhythm database. This paper presents a new method which extracts essential amplitude, duration and gradient parameterized features on processed ECG signal essential for human identification. Finally, bagged tree classifier (ensemble classifier) is utilized to evaluate the accuracy of our method. With this system, we obtained a high identification rate (97.5%)