A Technical Analysis and Practical Implementation of Machine Learning Algorithms for Predicting Survival in Breast Cancer Patients

Breast cancer is one of the most significant health challenges worldwide, with millions of women diagnosed annually and survival rates heavily dependent on early detection and effective treatment. The complexity of breast cancer, with its various subtypes and clinical characteristics, makes accurate survival prediction a critical yet challenging task. Machine learning algorithms have emerged as potent tools in predicting survival outcomes by analyzing extensive patient data and identifying complex patterns that may be overlooked by traditional clinical methods. While numerous studies have explored machine learning approaches for survival prediction in breast cancer patients, many lack comprehensive comparisons of different algorithms on specific datasets and practical implementation insights. This paper aims to advance the field by comparing the performance of seven prominent machine learning classification algorithms: Naive Bayes (NB), k-Nearest Neighbors (kNN), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), Logistic Regression (LR), and Artificial Neural Network (ANN). Utilizing the SEER (Surveillance, Epidemiology, and End Results) dataset from the National Cancer Institute, these classifiers are evaluated based on accuracy, precision, recall, and F-Score to determine their effectiveness in predicting patient survival. Among these algorithms, Random Forest shows superior performance, with an accuracy of 91.5%, precision of 92.80/0, recall of 55.7^%, and an F-Score of 69.6%, highlighting its robustness and efficacy in survival prediction tasks. While other algorithms also demonstrate potential, particularly under certain conditions, further research is needed to explore the benefits of combining multiple algorithms and incorporating additional features to enhance survival prediction models for breast cancer patients in future studies.