Reg2Boost: A Machine Learning Optimization Approach for Low-Dimensional Breast Cancer Classification

Early and accurate classification of malignant tumors is vital for patient diagnosis, treatment decisions, and prognosis evaluation. However, traditional diagnostic methods are subjective, have high misdiagnosis rates, and under-utilize data. This study presents the innovative Reg2Boost model for multi-algorithm collaborative optimization. It aims to solve the classification problem of low-dimensional data, especially take the Wisconsin Breast Cancer dataset as an example. The model uses K-means clustering, multiple linear regression, and an XGBoost classifier in sequence. Through data standardization, clustering, feature generation, model training, and hyperparameter tuning, it creates a systematic and efficient classification system. Experimental results show that Reg2Boost outperforms the traditional XGBoost model in key metrics like classification accuracy, recall rate, F1 score, and AUC. The Reg2Boost model is highly valuable. Clinically, it enables accurate breast cancer diagnosis for treatment planning. In research, it reveals key features, deepening our understanding of breast cancer biology. It also spurs the development of advanced medical machine-learning models with new algorithm combinations.