Scholarlink

Cancer detection is a critical medical process for identifying cancerous cells in the body. It plays a crucial role in early diagnosis, increasing the chances of effective treatment and improving patient survival. Machine learning (ML) techniques are transforming healthcare by enabling rapid and accurate medical data analysis. However, cancer detection remains a major challenge for practitioners and researchers. In this context, ensemble methods such as Bagging (Bgg), Random Forest (RF), and Boosting (Bst) have shown promise in addressing this challenge. This study aims to compare the performance of these approaches using the “Breast Cancer” dataset collected from the UCI machine learning repository. The algorithm is implemented in Python, and the following performances are noted: for RF, Accuracy is 0.96, AUC-ROC is 0.99; for Bgg, Accuracy is 0.96, AUC-ROC is 0.99; for Bst, Accuracy is 0.97, AUC-ROC is 1.00. This study will address the following question: How can ensemble methods overcome the complexity and imbalance of medical data in breast cancer detection, thereby surpassing the limitations of traditional models?

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Adaptation of Ensemble Methods to the Healthcare Field: Breast Cancer Detection

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mahmoudi Mohamed,
  • Azhari Mourad,
  • Ziti Soumia

摘要

Cancer detection is a critical medical process for identifying cancerous cells in the body. It plays a crucial role in early diagnosis, increasing the chances of effective treatment and improving patient survival. Machine learning (ML) techniques are transforming healthcare by enabling rapid and accurate medical data analysis. However, cancer detection remains a major challenge for practitioners and researchers. In this context, ensemble methods such as Bagging (Bgg), Random Forest (RF), and Boosting (Bst) have shown promise in addressing this challenge. This study aims to compare the performance of these approaches using the “Breast Cancer” dataset collected from the UCI machine learning repository. The algorithm is implemented in Python, and the following performances are noted: for RF, Accuracy is 0.96, AUC-ROC is 0.99; for Bgg, Accuracy is 0.96, AUC-ROC is 0.99; for Bst, Accuracy is 0.97, AUC-ROC is 1.00. This study will address the following question: How can ensemble methods overcome the complexity and imbalance of medical data in breast cancer detection, thereby surpassing the limitations of traditional models?