Risk factor identification mechanism for coronary artery disease based on multiple cross-filtering and binary cuckoo search.

Coronary artery disease (CAD) is a major public health concern, necessitating accurate risk factor identification. However, existing methods often suffer from feature preference bias and insufficient multidimensional evaluation, limiting their reliability. To address this, we propose a novel two-stage mechanism integrating multiple cross-filtering and binary cuckoo search (BCS). In the first stage, features are evaluated from three perspectives-relevance (chi-square test), information richness (mutual information), and distance (Fisher score)-to eliminate low-importance features and reduce bias. In the second stage, a random forest classifier optimizes feature selection via BCS, using classification accuracy as the objective function. Empirical analysis on the UCI CAD dataset demonstrates that our method achieves an accuracy of 89%, precision of 0.87, recall of 0.91, F1-score of 0.89, and AUC of 0.93. These values outperform existing homogeneous models (Method II, Method III, Method IV) by at least 3.49%, 3.57%, 3.41%, 3.49%, and 3.33%, respectively. The results highlight superior computational efficiency and predictive performance, making the mechanism a valuable tool for CAD risk assessment.