Comparative Analysis of Machine Learning Models for Rheumatoid Arthritis Prediction Using Multiclass Clinical Data
Comparative Analysis of Machine Learning Models for Rheumatoid ArthritisPrediction using Multiclass Clinical Data
Santosh Kalshetty
Department of Computer Engineering Siddant
Engineering College Pune, India
santoshkalshetty159@gmail.com
Prof. Nanda S. Kulkarni
Head of Department Siddant Engineering College Pune,
India
Abstract—Rheumatoid Arthritis (RA) is a chronic autoimmune inflammatory disorder that leads to joint destruction and long-term disability if not diagnosed early. Machine learning tech- niques enable efficient analysis of multidimensional clinical data for early prediction. This study presents a comparative evalu-tionofLogistic Regression, Support Vector Machine (SVM), Random Forest, and K-Nearest Neighbors (KNN) for multiclass arthritis classification. Performance is evaluated using accuracy, precision, recall, F1-score, confusion matrix, AUC-ROC curves, SHAP explainability, and statistical validation. Experimental resultsdemonstrate that Random Forest achieves superior overall performance.
Index Terms—Rheumatoid Arthritis, Multiclass Classification, Machine Learning, Random Forest, SHAP, ROC Curve