Breast Cancer Detection Using a PSO-ANN Machine Learning Technique

Abstract

Machine learning is employed in all facets of life. Breast cancer has been known to be the second most severe cancer that leads to death among women globally. The use of dimensionality reduction to reduce noise and eliminate irrelevant features from dataset is of enormous significant on breast cancer detection. In this study, particle swarm optimization (PSO) algorithm was employed to select relevant features from the data with artificial neural network for classification purpose on a University of California Irvine machine learning database dataset. The study was evaluated with the findings revealing the performance of the study at 97.13% accuracy. Conclusively, the aim of this study is to improve machine learning approach for breast cancer detection. This paper will be of help to radiologists in taking accurate results and making proper decisions regarding breast cancer early diagnosis based on machine learning.