Breast Cancer Prediction Using Hybridization of Machine Learning and Optimization Technique

A major global health concern is still breast cancer (BC) in today’s time, necessitating the development of precise and efficient prediction models to enhance early diagnosis and improve the results. The proposed work presents a new technique for prediction of BC by combining machine learning (ML) and optimization technique called genetic algorithm (GA). Two datasets, namely mammography mass and the BC dataset from UCI, were used in this work. Feature selection and preprocessing methods are implemented to increase the quality and relevance of input variables. The work focuses on developing a robust predictive model that helps identify individuals with the disease. A variety of ML algorithms are employed to develop the proposed predictive models. We evaluate model performance using accuracy, confusion matrix, support, and recall metrics. The highest accuracy we achieved using the proposed methodology is 86.14% on the mammography mass dataset.