Breast cancer detection using two-fold genetic evolution of neural network ensembles

Abstract

Breast cancer is the development of a malignant tumor notably in the breasts of a female. No proven cure is yet known for breast cancer, except when detected at an initial stage. This paper presents an innovative approach to the diagnosis of breast cancer by using two proposed variants of Genetic Algorithms, the Inter-Genetic Algorithm, and the Intra-Genetic Algorithm, that evolves an ensemble of Neural Networks and its constituent Artificial Neural Networks, respectively. The proposed approach obtains an average accuracy of 99.90% using 70–30% training to testing ratio on the Wisconsin Breast Cancer dataset and hence is a reliable alternative for providing a second opinion to human experts for the classification of breast cancer tumors.