•  
  •  
 

Abstract

The particle swarm optimization (PSO) algorithm was applied to penalize the Cox model for predicting long-term outcomes in breast cancer patients. This study utilized data from 198 breast cancer survivors,including their age, estrogen receptor status, tumor size, grade, and expressionlevels of 76 genes. The aim was to identify a feature subset that could accurately predict patient survival while mitigatingoverfitting and model complexity. PSO was used to search for optimal model parameters. The algorithm was designed to minimize a penalized partial likelihood function, which balances the trade-off between model accuracy and complexity.The values of the objective function were compared with that of other feature selection techniques, including the least absolute shrinkage and selection operator (LASSO) and elastic regression,and found to outperform them in predictive accuracy and feature selection. Results demonstratedthat PSO-PCOXwith cross-validation forregularization parameters achieved higher prediction accuracy than models trained withother feature selection methods. The PSO algorithm identified a subset of features that were consistently selected across multiple iterations, indicating their importance in predicting patient survival. Overall, thisstudy showcases the potential of PSO-based feature selection in enhancing the accuracy and interpretability of Cox regression models for predicting long-term outcomes in breast cancer patients

Download

Share

COinS