Optimized Machine Learning Approach for Predicting Venous Clinical Severity Scale Outcomes in Yoga Practitioners

Yoga is recognized as one of the oldest holistic health systems originating in India and has gained increasing scientific attention for its therapeutic potential in managing chronic vascular and circulatory disorders. The objective of this study is to develop an optimized machine learning–based predictive framework for estimating Venous Clinical Severity Scale (VCSS) outcomes in individuals practicing yoga, with the aim of evaluating yoga’s effectiveness in the management of varicose veins. To enhance classification performance through efficient hyperparameter tuning, two advanced metaheuristic optimization algorithms—Prairie Dog Optimization (PDO) and Coronavirus Herd Immunity Optimizer (CHIO)—were integrated with a k-Nearest Neighbor Classification (KNNC) model. Disease severity was categorized into four clinically relevant levels, namely absent, mild, moderate, and severe, using both VCSS-pre and VCSS-1 indices as diagnostic indicators. Model performance was systematically evaluated using standard classification metrics, including accuracy, precision, recall, and F1-score. The experimental findings demonstrate that the PDO-optimized model (KNPD) achieved the highest predictive performance, improving classification accuracy by 2.38% for VCSS-pre and 2.27% for VCSS-1 compared with the baseline KNNC model. The CHIO-optimized model (KNCH) also exhibited measurable improvements in predictive accuracy, although its performance gains were comparatively lower than those of KNPD. Overall, the KNPD framework showed superior classification capability, reduced misclassification rates, and enhanced reliability. These results indicate that the proposed optimized learning model provides a robust and effective computational tool for assessing venous disease severity and offers strong evidence supporting the potential role of yoga as a complementary therapeutic intervention in the treatment and management of varicose veins.