Scholarlink

<p>Recently, the growing population has increased the challenge of early disease detection. Breast cancer is the second most severe cancer, and early diagnosis is critical for reducing mortality. This study introduces a novel hybrid machine learning approach for breast cancer prediction, integrating Agglomerative Hierarchical Clustering (AHC) with multiple classifiers such as Gradient Boosting (GB), Stochastic Gradient Descent (SGD), Quadratic Discriminant Analysis (QDA), and Naive Bayes (NB). This integration leverages clustering to group similar instances, improving classifier focus and predictive accuracy. The model’s performance is evaluated across multiple metrics including accuracy, precision, recall, F1-score, sensitivity, specificity, and false-positive/negative rates. The proposed approach demonstrates high effectiveness, with AHC + GB achieving the highest accuracy of 99.30%, outperforming standalone methods. This research highlights the novelty of combining unsupervised and supervised techniques, identifies research gaps, and provides a foundation for clinically relevant, robust breast cancer prediction systems.</p>

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

Innovative Hybrid Machine Learning Methods for Breast Cancer Predictive Optimization

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月17日

查看全文
  • Muhammad Zohaib Khan,
  • Muqaddas Salahuddin,
  • Abdullah Ayub Khan,
  • Syed Akmal Sultan,
  • Shehzeen Dua Bhatti,
  • Sajid Ullah,
  • Mohamad Afendee Mohamed

摘要

Recently, the growing population has increased the challenge of early disease detection. Breast cancer is the second most severe cancer, and early diagnosis is critical for reducing mortality. This study introduces a novel hybrid machine learning approach for breast cancer prediction, integrating Agglomerative Hierarchical Clustering (AHC) with multiple classifiers such as Gradient Boosting (GB), Stochastic Gradient Descent (SGD), Quadratic Discriminant Analysis (QDA), and Naive Bayes (NB). This integration leverages clustering to group similar instances, improving classifier focus and predictive accuracy. The model’s performance is evaluated across multiple metrics including accuracy, precision, recall, F1-score, sensitivity, specificity, and false-positive/negative rates. The proposed approach demonstrates high effectiveness, with AHC + GB achieving the highest accuracy of 99.30%, outperforming standalone methods. This research highlights the novelty of combining unsupervised and supervised techniques, identifies research gaps, and provides a foundation for clinically relevant, robust breast cancer prediction systems.