Breast cancer represents a significant global health challenge, underlining the necessity of early detection to enhance patient outcomes. This study examines the effectiveness of three machine learning approaches, Decision Tree, Random Forest, and Gradient Boosting in identifying breast cancer using a comprehensive dataset that includes demographic and clinical attributes. The Decision Tree technique provides interpretability by dividing data based on feature values. Random Forest, an ensemble method, enhances performance by combining multiple decision trees to mitigate overfitting and improve classification accuracy. Gradient Boosting, another ensemble method, incrementally improves predictions by correcting previous errors, resulting in superior accuracy. Metrics such as accuracy, precision, recall, and F1-score are utilized to evaluate the models, demonstrating that Gradient Boosting excels in accuracy and robustness, while Random Forest provides balanced performance. These methods illustrate the potential of machine learning to enhance breast cancer diagnosis, enabling more precise and efficient detection.