Detection of early breast cancer using a-priori rule mining and machine learning approaches

Abstract

In today’s world, breast cancer is extremely predominant in females that establishes in the breast and further extends to other locales of the body in the track of time. It is the second major ailment that causes decease. In long term, an early detection can reduce the death rate due to breast cancer appreciably. The crucial point for early prediction is to recognize the cancer cells at virgin stages. Various researches are carried out on breast cancer detection using mammography, ultrasounds, CT scans, PET, MRI, biopsy, etc. Still, these techniques are expensive, prolonged and sometimes unsuitable for young females. Hence, a fast and accurate detection system is highly demanded. In recent years, data mining and machine learning techniques are given utmost attention for early stage breast cancer detection. The aim of this paper is to present a framework for accurate and quick conclusion of breast cancer using machine learning techniques. We applied our proposed technique on SEER dataset of breast cancer and obtained highly appreciable results with accuracy of 99.9% using random forest. Various rules are also presented in support of breast cancer detection using A-priori algorithm.