Genetic Algorithm Based Selection of Appropriate Biomarkers for Improved Breast Cancer Prediction

One of the most common types of cancer among women is Breast Cancer which amounts to a staggeringly high number of deaths every year. According to the World Health Organization (WHO), the projected number of Breast Cancer cases in 2018 is 2.09 million. However, early diagnosis of Breast Cancer does lead to a much higher five year survival rate. In this study, Genetic Algorithm (GA) is used to identify appropriate biomarkers for predicting Breast Cancer, from routine consultation and blood analysis data, which can subsequently help in the process of early diagnosis of the disease. It was observed during experimental study, that Resistin, Leptin, Glucose, BMI and Age of the patients, can be used for the purpose of automatic breast cancer prediction. In this paper, Breast Cancer Coimbra Dataset (BCCD) has been considered for performing experimentations. The GA-selected patient-attributes were subsequently utilized, by a number of Machine Learning (ML) classifiers, to learn a prediction model for the disease. It has been observed that, GA combined with Gradient Boosting Classifier performed better than all other classifiers under consideration, in terms of prediction accuracy, AUC and F-measure values, suggesting the potential applicability of this model in practical scenarios.