A Hybrid Supervised Machine Learning Classifier System for Breast Cancer Prognosis Using Feature Selection and Data Imbalance Handling Approaches

Nowadays, breast cancer is the most frequent among women. Early detection a critical issue that can be effectively achieved by machine learning (ML) techniques. Thus in this article, methods to improve accuracy of ML classification models for prognosis are investigated. Wrapper-based feature selection approach along with nature-inspired algorithms such as Particle Swarm Optimization, Genetic Search, and Greedy Stepwise has been used identify important features. On these selected features popular classifiers Support Vector Machine, J48 (C4.5 Decision Tree Algorithm), Multilayer-Perceptron (a feed-forward ANN) were system. The methodology proposed system structured into five stages which include (1) Data Pre-processing; (2) imbalance handling; (3) Feature Selection; (4) Machine Learning Classifiers; (5) classifier’s performance evaluation. dataset under research experimentation referred from UCI Repository, named Breast Cancer Wisconsin (Diagnostic) Set. This article indicated decision tree classifier appropriate learning-based optimum prognosis. Optimization algorithm achieves 98.24%, MCC = 0.961, Sensitivity 99.11%, Specificity 96.54%, Kappa statistics 0.9606. It also observed Search 98.83%, 0.974, 98.95%, 98.58%, 0.9735. Furthermore, Multilayer Perceptron ANN 98.59%, 0.968, 98.6%, 98.57%, 0.9682.