Authors:
Zeyneb Kurt
and
Oguzhan Yavuz
Affiliation:
Yildiz Technical University, Turkey
Keyword(s):
Auto-regressive Model, HIV, Support Vector Machine, ROC Analysis.
Related
Ontology
Subjects/Areas/Topics:
Biometrics and Pattern Recognition
;
Multimedia
;
Multimedia Signal Processing
;
Neural Networks, Spiking Systems, Genetic Algorithms and Fuzzy Logic
;
Telecommunications
Abstract:
We propose intelligent methods to classify two different HIV virus types, i.e., R5X4 and R5 or X4 with low computational complexity. Since R5X5 virus has same the features of R5 and X4 viruses, diagnosis of R5X4 can not be determined easily. In this study, the statistical data of R5X4, R5 and X4 was obtained by accessible residues and modelled by Auto-regressive (AR) model. After that the pre-processed data was used for determining the optimal value in Radial Basis Kernel of Support Vector Machine (SVM).