Friday 24th of November 2017
 

Knee Osteoarthritis Diagnosis Using Support Vector Machine and Probabilistic Neural Network


Ozge Aksehirli, Duygu Aydin, Handan Ankarali and Melek Sezgin

Support Vector Machines (SVM) and Probabilistic Neural Networks (PNN) that are within the scope of artificial intelligence applications has been widely accepted. Because of strong statistical foundations of these methods, they generate generalizable results. In this study we aimed to compare classification performance of SVM and PNN methods using real and empirical data sets. In the real data set, we investigated the success of diagnosis of some demographic characteristics and some gene polymorphisms on knee osteoarthritis. In addition, success of SVM and PNN methods were compared for classification of samples that have various data structure. We found that the classification performance of PNN method is better than SVM. The most important disadvantage of PNN method is that the analysis time increase with sample size and number of attributes. When the proportion of the observation number to attributes number is greater than 5, SVM and PNN can be used for classification.

Keywords: Optimization, classification, radial basis function, support vector machine, neural networks

Download Full-Text


ABOUT THE AUTHORS

Ozge Aksehirli
Department of Biostatistics and Medical Informatics, Faculty of Medicine, University of Düzce Düzce, 81000,Turkey

Duygu Aydin
Department of Biostatistics, Faculty of Medicine, University of Hacettepe Ankara, 06000, Turkey

Handan Ankarali
Department of Biostatistics and Medical Informatics, Faculty of Medicine, University of Düzce Düzce, 81000,Turkey

Melek Sezgin
Department of Physical Therapy and Rehabilitation, Faculty of Medicine, University of Mersin Mersin, 33000, Turkey


IJCSI Published Papers Indexed By:

 

 

 

 
About IJCSI

IJCSI is a refereed open access international journal for scientific papers dealing in all areas of computer science research...

Learn more »
Join Us
FAQs

Read the most frequently asked questions about IJCSI.

Frequently Asked Questions (FAQs) »
Get in touch

Phone: +230 911 5482
Email: info@ijcsi.org

More contact details »