Pemilihan Fitur Menggunakan Metode Ensemble Lasso Regression, Random Forest dan Recursive Features Elimination Dalam Klasifikasi Kanker Payudara

Abstract

Healthcare datasets, especially those used in cancer diagnosis, often present challenges such as high dimensionality, redundancy, and irrelevant features, which can reduce the performance and reliability of automated learning models. This study proposes a robust ensemble feature selection method to address these challenges, by combining Lasso Regression, Random Forest, and Recursive Feature Elimination (RFE). By utilizing the complementary strengths of these algorithms, the ensemble approach aims to improve the stability of feature selection and enhance classification accuracy. In addition, Shannon entropy is used to evaluate data complexity and guide the feature selection process. The proposed method is applied to the Breast Cancer (Diagnosis) dataset and its performance is evaluated using metrics such as accuracy, precision, gain, and F1 score. The experimental results show that the ensemble method outperforms individual feature selection techniques, achieving higher classification accuracy and reliability in handling complex and imbalanced datasets. This research advances machine learning-based diagnostic tools by providing a reliable framework for analyzing high dimensional medical data. These results highlight the potential of synthetic feature selection to improve interpretation, reduce computational cost, and increase the predictive accuracy of breast cancer diagnosis, revealing the potential of synthetic feature selection.