Published in

American Scientific Publishers, Journal of Computational and Theoretical Nanoscience, 2(16), p. 627-632, 2019

DOI: 10.1166/jctn.2019.7781

Links

Tools

Export citation

Search in Google Scholar

Evaluating the Efficiency of Radial Basis Function Classifier with Different Feature Selection for Identifying Dementia

Journal article published in 2019 by S. Valarmathy, R. Ramani
This paper was not found in any repository, but could be made available legally by the author.
This paper was not found in any repository, but could be made available legally by the author.

Full text: Unavailable

Red circle
Preprint: archiving forbidden
Red circle
Postprint: archiving forbidden
Red circle
Published version: archiving forbidden
Data provided by SHERPA/RoMEO

Abstract

The Magnetic Resonance Imaging (MRI) based classification process for the classification of dementia is presented in this work. The classifier's performance may be enhanced by means of improving the extracted features that are inputted into its classifier. These MRI images are all duly segmented by making use of the wavelet. For choosing a subset that has optimal features, it may become inflexible and all issues relating to the feature selection will be shown as the NonDeterministic Polynomial (NP)-hard. The work further deals with techniques of optimization that are used in the case of feature selection for picking an optimal feature set. The Principal Component Analysis (PCA) will find an application of a large scale in signal processing. The noise estimation and the source separation are all possible. For this, the Radial Basis Function (RBF) and its classifier have been optimized to this structure by making use of the Genetic Algorithm (GA)-Artificial Immune System (AIS) algorithm. Such an optimized classifier of the RBF will classify a feature set that is provided by the GA, the AIS and the GA-AIS algorithm of feature selection. A classifier will be evaluated on the basis of its performance metrics. All classifiers will be evaluated keeping the accuracy, specificity, and sensitivity in making use of an optimized set of features. The results of the experiment have clearly demonstrated the feature selection and its effectiveness to improve the accuracy of the classification of all the images.