Multivariate Analysis of EEG data using Fractal Dimension
*Md Rokonuzzaman, Student
Keywords: Discriminant Analysis, Electroencephalogram, Fractal Dimension, Mahalanobis distance, MANOVA, Max-eigen difference method, Outlier.
The goal of this study is to work on diagnostic of the multivariate model. The main focus was on multivariate analysis of EEG data which comprised multivariate analysis of variance and discriminant analysis. The multivariate model diagnostics comprise checking number of assumptions (multivariate normality, homogeneity of covariance matrices) and presence of outliers. Since multivariate analysis of variance is extremely sensitive to the presence of outliers, the main interest of this study is to investigate the performance of a newly proposed method (Max-eigen difference method) for identification of multivariate outliers based on the eigenvalues of the sample covariance matrix. The obtained results from an empirical study showed that Max-eigen difference method to identify multivariate outliers’ works better compared to Mahalanobis distance method.
Important Dates & Deadlines
- October 9 - 11, 2013