Hybrid Machine Learning Model for Breast Cancer Detection

Abstract

Breast cancer is considered one of the major reasons for mortality in the female population around the world. It is highly important for medical practitioners to detect breast cancer as early as possible in order to provide adequate treatment and increase the chances for patients' survival. Thus, in the current study, the machine learning-based technique will be used for the prediction of breast cancer at an early stage using the Wisconsin Breast Cancer Dataset. It should be noted that the dataset consists of a set of different diagnostic features obtained from digitized image of fine needle aspirate (FNA) of breast mass. The supervised algorithms, including K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Random Forest, will be applied for the classification of the dataset into benign and malignant tumors. Pre-processing procedures are also implemented in the experiment in order to improve the effectiveness of machine learning algorithms. These procedures involve data cleansing, normalization, and feature selection. Performance of different models is evaluated on the basis of such metrics as accuracy, precision, recall, and F1-score. According to experimental data, Random Forest classifier demonstrated the highest accuracy.