Maria Miroiu

In this article, we describe a new classification methodology based on the use of Independent Component Analysis and Wavelet decomposition (ICAW) techniques. An ensemble system of classifiers is built such that each classifier independently decides the assignation of the test examples on several representations resulted by taking projections computed by wavelets and Independent Component Analysis (ICA). The representations used by the individual classifiers are obtained by taking the real and imaginary part of the wavelet decompositions, as well as the magnitude and phase. The decision of the ensemble system is based on several types of voting rules (such as the majority voting rule or a weighted voting rule). The experimental results presented in the paper show that the proposed ensemble systems of classifiers provide higher accuracy in the particular problem of classifying biomedical data.

The aim of the paper is to propose a new methodology for solving classification tasks based on ICAW (Independent Component Analysis and Wavelets) . The general idea is to use an ensemble system of classifiers that decide independently on representations resulted by taking projections computed by wavelets and Independent Component Analysis (ICA) and to combine the decisions of the particular classifiers in order to improve the overall decision. Each classifier establishes its decision on a different projection of the data and for each individual combined with an ensemble system of classifiers, the best classification for any individual being computed by combining the decisions of the particular considered classifiers. Each classifier uses a different projection of the data. The individual classifiers perform classification on real and imaginary coefficients, magnitude and phase corresponding to the representations of the data. The ensemble system of classifiers essentially based on I...