This study investigates the application of supervised machine learning for classifying breast tumors as benign or malignant, leveraging the Breast Cancer Wisconsin Dataset. The proposed method encompasses a comprehensive pipeline, beginning with data preprocessing to address missing values and ensure feature normalization. Exploratory Data Analysis (EDA) techniques are employed to uncover patterns and relationships within the data. To enhance model performance, feature selection is performed using various techniques, including correlation-based selection, tree-based methods, and Recursive Feature Elimination with Cross-Validation (RFECV). Machine learning algorithms, including Random Forest (RF), SVM, Logistic Regression (LR), and Gradient Boosting (GB), were trained on the selected features. Hyperparameter tuning was performed using grid and randomized search to optimize model accuracy. The results demonstrate the effectiveness of the proposed method, achieving significant improvements in classification metrics such as precision, recall, F1-score, and ROC-AUC. These findings underscore the potential of machine learning to enhance diagnostic accuracy and reliability, offering a scalable, efficient, and robust approach to breast cancer diagnosis. This work paves the way for future integration of advanced techniques, including deep learning models and larger datasets, to further improve diagnostic outcomes and accessibility.
Naganandini et al. (Sat,) studied this question.