Application of SVM with Python and Machine Learning for Effective Breast Cancer Detection

Melgarejo-Solis R.; Roman-Concha U.; Romero-Untiveros L.; Arboleda-Sanchez E.; Yucra-Tintaya A.; Figueroa-Estrella J.; Zavala-Tapia A.; Soto-Vargas L.; Isusqui J.P.; Luera-Collazos J.; Herrera-Chavez C. · Advances in Artificial Intelligence and Machine Learning (2025) · DOI: 10.54364/aaiml.2025.53244

Resumen

Breast cancer remains one of the leading causes of female mortality worldwide, highlighting the need for early diagnostic tools to support medical work. This study evaluated the application of machine learning algorithms for breast tumor classification using the public Breast Cancer Wisconsin (Diagnostic Dataset) dataset. The methodological process included variable cleaning and normalization, correlation analysis, and feature selection, followed by the implementation of a Support Vector Machine (SVM) model with a Radial Basis Function (RBF) kernel. Hyperparameters were optimized using grid search, and performance was validated with a k=10 cross-validation scheme. The results showed that the SVM achieved an accuracy of 96%, with a sensitivity of 95% and a specificity of 97%, outperforming reference models such as logistic regression (91%), decision trees (88%), and KNN (90%). These findings confirm the ability of SVMs to model nonlinear relationships in high-dimensional biomedical data. Furthermore, the model was integrated into a Django-based web prototype, which allows image uploads and the retrieval of probabilistic diagnoses in real time, demonstrating its potential for practical application in hospital settings. The results show that SVM with an RBF kernel constitutes an alternative for the development of support systems for the early diagnosis of breast cancer.

Instituciones: Universidad Nacional Mayor de San Marcos; Universidad Nacional Federico Villarreal; Universidad Privada del Norte