Effective Prediction of Vertebral Column Pathologies Using Ensemble Classifiers

Authors

  • K. N. Nithya
  • P. Suresh

Abstract

The medical diagnosis had a greater impact in the medical research domain. Accurate classification plays a vital role in medical diagnosis. It avoids the complexities and enables the treatment stage very effective. This study develops a prominent method that results in accurate recognition of the vertebral column pathologies. In our previous work, we developed a SPRINT algorithm which is a single classifier. Using single classifier counter parts, the possibility of poor selection. To overcome this and enhance the classification performance, we propose multiple-classifier techniques with multiple voting model. In this paper, we develop an Ensemble classifier for processing classification. The ensemble classifier applies to label the vertebral disorder image based on the similarity features. Ensemble classifier has a certain set of classifiers each classifier creates its model and combines response taken for achieving excellence in classification performance. The experimental work on this research is carried out with MATLAB and WEKA tools using UCI medical dataset. The performance evaluation based on the obtained results is achieved by undergoing several evaluation metrics. The efficiency of classification is measured by sensitivity, specificity, F-measure, and accuracy. The obtained result ensures our implementation of the ensemble classifier achieves better accuracy in classification and classifier speeds compared to others..

Downloads

Published

2020-01-30

Issue

Section

Articles