Title: Hybrid Support Vector Machine for imbalanced data in multiclass arrhythmia classification

Authors: Aniruddha J. Joshi, Sharat Chandran, V.K. Jayaraman, B.D. Kulkarni

Addresses: ViGIL, Department of Computer Science and Engineering, Indian Institute of Technology Bombay, Mumbai 400076, India. ' ViGIL, Department of Computer Science and Engineering, Indian Institute of Technology Bombay, Mumbai 400076, India. ' Chemical Engineering and Process Development Division, National Chemical Laboratory, Pune 411008, India. ' Chemical Engineering and Process Development Division, National Chemical Laboratory, Pune 411008, India

Abstract: Automatically classifying ECG recordings for arrhythmia is difficult since even normal ECG signals exhibit irregularities, and learning algorithms suffer from class imbalance. We propose a hybrid SVM to combat class imbalance rampant in biomedical signals. Consequently, we significantly reduce the number patients falsely classified as normal. The Hybrid SVM is suitable for a variety of multiclass problems; here, we used the MIT-BIH Arrhythmia database, and the position and magnitude of local singularities as features. We enhance relevant singularity-driven Holder features proposed earlier; while the use of these features results in higher accuracy, using the Hybrid SVM shows even more gains.

Keywords: multiclass classification; SVM; support vector machine; imbalanced data; local Holder exponents; class imbalance; local singularity; hybrid SVM; arrhythmia; personalised medicine; ECG signals; electrocardiograms.

DOI: 10.1504/IJFIPM.2010.033244

International Journal of Functional Informatics and Personalised Medicine, 2010 Vol.3 No.1, pp.29 - 47

Published online: 14 May 2010 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article