Dimension and Complexity Study for Alzheimer’s disease Feature Extraction

Abstract

This paper discusses the problem of dimensionality and computational complexity analysis for feature extraction proposed algorithms of Alzheimer’s disease. The effective features are very useful for some of the discrimations and to assist the physicians in the detection of abnormalities. This paper concern two main issues that must be confronted which are: The first one concern the study of how the classification accuracy depends on the dimensionality (i.e. the number of features). The second issue is the computational complexity of designing the classifier. As the number of features increases, the classification error decreases which consequently improve the accuracy of the classifier.