<![CDATA[This study developed a multi-algorithm machine learning prototype for multiclass breast cancer survival prediction using 1,980 patient records, classifying patients as Living, Died of Disease, or Died of Other Causes. The framework integrated NN, SVM, RF, NB, and KNN algorithms within a decision-support monitoring application, with preprocessing steps including data cleaning, normalization, feature preparation, and dataset partitioning. To prevent target leakage, survival-related variables were excluded from the predictor set. The revised evaluation results indicated that NB and KNN delivered the strongest performance, achieving weighted average F1-scores of 0.93 and 0.92, respectively, while NN and RF showed comparatively lower results. These findings highlight the potential of machine learning for breast cancer survival status monitoring, although the proposed system is designed as a decision-support prototype rather than a clinical diagnostic tool. Therefore, before actual healthcare deployment, more research incorporating explainable AI techniques, external validation, and real-world clinical testing is needed.]]>
Copyrights © 2026
