Improved feature selection and classification for rheumatoid arthritis disease using weighted decision tree approach (REACT)

Rheumatoid arthritis (RA) is a major chronic autoimmune disorder which affects multiple organs and joints of human body. Disease varies in its behavior and concern such that an early prediction is a complex process with regard to time so the diagnosis is not an easy task for the physicians. The comm...

Full description

Saved in:
Bibliographic Details
Published in:The Journal of supercomputing 2019-08, Vol.75 (8), p.5507-5519
Main Authors: Shanmugam, S., Preethi, J.
Format: Article
Language:English
Subjects:
Arthritis
Autoimmune diseases
Classification
Compilers
Computer Science
Data mining
Decision trees
Diagnosis
Human behavior
Interpreters
Organs
Particle swarm optimization
Physical examinations
Physicians
Predictions
Processor Architectures
Programming Languages
Rheumatoid arthritis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Rheumatoid arthritis (RA) is a major chronic autoimmune disorder which affects multiple organs and joints of human body. Disease varies in its behavior and concern such that an early prediction is a complex process with regard to time so the diagnosis is not an easy task for the physicians. The common existing methodologies employed to analyze the severity of RA are the clinical, laboratory and physical examinations. The advancement of data mining has been employed for the RA diagnosis through learning from history of datasets. To improve the efficiency and reliability of the approach, this paper presents a hybrid optimization strategy called REACT, which is based on the combination of the features of Iterative Dichotomiser 3 and Particle Swarm Optimization for feature selection and classification of RA. The effectiveness of the proposed diagnosis strategy is validated through its prediction accuracy, specificity, sensitivity, positive predictive value and negative predictive value with existing approaches.
ISSN:0920-8542
1573-0484
DOI:10.1007/s11227-019-02800-1