Full Program »
Impact of Parameter Tuning on Machine Learning based Breast Cancer Classification
Breast cancer is one of the major causes of death among women. Different decision support systems were proposed to assist oncologists in order to accurately diagnose their patients. These decision support systems mainly used Machine Learning (ML) techniques to classify the diagnosis into malign or benign tumor. In this paper, we evaluate and analyze the accuracy of the parameters tuning on the accuracy of three well-known ML techniques: Support Vector Machines (SVM), Multi-Layer Perception (MLP) and Decision trees (DTs). We investigate three parameters tuning techniques: Grid Search (GS), Particle Swarm Optimization (PSO) and the default strategy of the Weka Tool (Uniform Configuration, UC-Weka)) over four datasets obtained from the Machine Learning repository. The overall results suggest that using GS and PSO lead to build more accurate classifiers, and therefore can help oncologists to provide more accurate diagnosis.