Comparative analysis of similarity measures for medical streaming data

Abstract

In this paper, we analyse the similarity measures for comparison of medical streaming data (MSD). In general, the real time streaming data can be inter- preted as the multivariate time series. In medicine, the comparison of patient physiological MSD data can be useful in disease detection. Four similarity mea- sures Correlation Coefficient, Frobenius norm, Principal Component Analysis similarity factor, Multidimensional Dynamic Time Warping are presented and the experiments of the comparison of these measures have been performed.