Optimizing Adaptive Boosting Model for Breast Cancer Prediction Using Principal Component Analysis and Random Oversampling Techniques

Breast cancer is the most common type of cancer in women and remains the leading cause of cancer death among them. Risk factors such as obesity, lack of physical activity, alcohol consumption, hormone therapy during menopause, radiation exposure, and family history plays important roles in its development. Early detection is critical, but machine learning applications for prediction face challenges, particularly due to class imbalance in the dataset, which can seriously impact model performance. This study focuses on optimizing AdaBoost parameters using a combination of PCA and Random Oversampling. The results show that the optimized model achieves 98.24% accuracy in breast cancer prediction. The combination of PCA for feature reduction and Random Oversampling for data balancing effectively improves prediction accuracy. These findings provide a solid foundation for developing more precise diagnostic methods using machine learning in future breast cancer research.