CLASSIFICATION AND DETECTION OF BREAST CANCER USING MACHINE LEARNING ALGORITHMS

Early breast cancer detection is crucial for improving patient outcomes, yet current diagnostic methods often lack precision and interpretability. This study introduces an innovative two-stage machine learning approach using Random Forest and Logistic Regression algorithms to enhance breast cancer classification and risk assessment respectively. Utilizing the Wisconsin Breast Cancer Dataset, the research develops a comprehensive methodology that addresses key challenges in medical diagnostics. The first stage employs Random Forest for robust feature extraction and selection, effectively reducing dimensionality while maintaining critical predictive information. The second stage utilizes logistic regression which provided a probabilistic framework for risk assessment based on the selected features, allowing clear interpretation of factors that influences breast cancer risk. Evaluation metrics such as accuracy, precision, recall and F1-score are employed to validate the model’s performance, with results indicating high accuracy (95.88%) and robustness against overfitting. The study concludes that this dual-model approach enhances the interpretability and reliability of breast cancer detection models, paving the way for clinical applications that support early diagnosis and targeted treatment strategies. This research highlights the potential of integrating machine learning in healthcare for early cancer detection, personalized risk assessments and ultimately improved patients’ outcomes.