文件列表:

embd.m (530, 2011-12-28)
__MACOSX (0, 2011-12-31)
__MACOSX\._embd.m (205, 2011-12-28)
mmse.m (4070, 2011-12-29)
__MACOSX\._mmse.m (205, 2011-12-29)
mmse_correlated_vs_uncorrelated.m (4279, 2011-12-28)
__MACOSX\._mmse_correlated_vs_uncorrelated.m (205, 2011-12-28)
mvsampen_full.m (1383, 2011-12-28)
__MACOSX\._mvsampen_full.m (205, 2011-12-28)
mvsampen_naive.m (1300, 2011-12-28)
__MACOSX\._mvsampen_naive.m (205, 2011-12-28)
powernoise.m (1952, 2009-10-27)
__MACOSX\._powernoise.m (205, 2009-10-27)

The following contains the description of the m-files available in the toolbox for Multivariate Multiscale Complexity Analysis 1. Main Matlab script: Filename: mmse.m - This script generates multi-channel white and 1/f noise and calculates multivariate multiscale sample entropy estimates over different temporal scales. It is used to generate Figure 2 of Ref [1] below. - The script produces a graph of MMSE estimates over different temporal scales (using coarse graining), giving an assessment of the complexity of the underlying dynamical structures in the data. - The method uses the mvsampen_full.m function to generate full multivariate sample entropy for multichannel data over different temporal scales. - Alternatively, you can replace this script with your own multivariate entropy estimates, for instance the 'naive' approach described below. - To analyse MMSE of your own general data, please replace the corresponding lines in the script with your own data realisations. 2. Supporting Matlab scripts: If you would like to look into the comparison of the 'naive' and 'full multivariate' approach to calculate MSampEn or into the ability of the 'full multivariate' approach to discriminate between correlated and uncorrelated white as well as 1/f noise, the following Matlab programs are available: a) Filename: mmse_correlated_vs_uncorrelated.m - This code calculates multivariate sample entropy over different scales of bivariate correlated and uncorrelated 1/f as well as white noise. It was used to generate Figure 3 of Ref [2] below. b) Filename: mvsampen_naive.m - This function is used to calculate MSampEn using the 'naive' multivariate approach, where the cross-channel correlations are not taken into account. This method was used in Reference [4] below. c) Filename: mvsampen_full.m - This function is used to calculate multivariate sample entropy (MSampEn) using the full multivariate approach, where both the inter-channel and cross-channel dependencies are taken into account; this method is used in References [1] and [2] below. d) Filename: embd.m - This function creates multivariate delay embedded vectors with the Embedding vector parameter M and time lag vector parameter tau. e) Filename: powernoise.m - This function generates samples of power law noise. The power spectrum of this signal scales as f^(-alpha). This function is written by Max Little [Ref. 3], and was downloaded from http://www.maxlittle.net/software/index.php. References: [1] M. U. Ahmed and D. P. Mandic, "Multivariate multiscale entropy: A tool for complexity analysis of multichannel data," Physical Review E, vol. 84, no. 6, pp. 061918-1 – 061918-10, 2011. [2] M. U. Ahmed and D. P. Mandic, "Multivariate multiscale entropy analysis," IEEE Signal Processing Letters, in press, 2012. [3] M.A. Little, P.E. McSharry, S.J. Roberts, D.A.E. Costello, I.M. Moroz (2007), "Exploiting nonlinear recurrence and fractal scaling properties for voice disorder detection", BioMedical Engineering OnLine 2007, 6:23. [4] M. U. Ahmed, L. Li, J. Cao, and D. P. Mandic, "Multivariate multiscale entropy for brain consciousness analysis", Proceedings of the IEEE EMBC Conference, pp. 810-813, 2011.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：