Optimized Breast Cancer Prediction Using Machine Learning Algorithms

Breast cancer remains a leading cause of cancer-related mortality among women worldwide, underscoring the critical need for early detection and accurate prediction to improve treatment outcomes. This study optimizes breast cancer prediction using various machine learning algorithms, including K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Trees, and ensemble methods such as Bagging and Boosting. The Wisconsin Diagnostic Breast Cancer (WDBC) dataset was employed, which was preprocessed to handle missing values and reduce dimensionality. Results indicate that the Decision Tree classifier achieved the highest accuracy at 98%, followed closely by KNN at 97.86%. KNN’s success can be attributed to its local approach, considering the five nearest neighbors, enabling it to adapt well to dataset variations. SVM, with an accuracy of 96.43%, excels in high-dimensional spaces by constructing optimal hyperplanes. Ensemble methods, particularly Bagging, reached 97.1%, demonstrating potential for stability through aggregation of multiple models. Despite ensemble techniques generally enhancing performance, they did not surpass the top individual classifiers in this analysis. These findings highlight the promise of machine learning in early breast cancer detection and the importance of selecting algorithms suited to specific dataset characteristics. Future research may focus on integrating advanced algorithms and larger datasets to refine predictive accuracy.