Breast Cancer Prediction Using Data Mining Classification Techniques

In today's world Breast Cancer has become the major source of mortality among women especially in underdeveloped countries like Pakistan, Sri Lanka, and Bangladesh. This is a highly alarming situation and needs the attention of the research community as there are not enough resources and health facilities. The rate of incidence could be reduced if the cancer is diagnosed at an early stage instead of late stages. Breast cancer occurs when some breast cells begin to rise abnormally. This research study intends to predict breast cancer by analyzing a set of attributes that have been selected from several classifications so that prevention can be done in time before it becomes incurable. This research work focuses on different classification techniques of data mining to predict Breast Cancer such as Decision Trees, Random Forest, Logistic Regression, Support Vector Machine, and Linear Discriminant Analysis and their comparative analysis for accurate disease detection. The dataset of breast cancer histopathology images was acquired from online recourses consisting of 277,524 images. The experimental results show that Random Forest performs better than all other algorithms used in this research study with an accuracy of 88.80 %, precision of 83.71 %, and recall 94.28 %. Python programming language to implement and perform the comparative analysis of algorithms used in this research work.