Breast cancer data analysis using support vector machines and particle swarm optimization

We propose a machine learning method for breast cancer data analysis and classification, based on support vector machines (SVM) and particle swarm optimization (PSO). This method uses SVM as a model for supervised learning with the goal of minimizing generalization errors, and PSO as an optimization technique for automatic determination of the best values of two algorithmic parameters of SVM. Its performance in solving classification and recognition problems is experimentally tested for a real-world benchmark dataset. The experimental results are compared to those provided by four other methods using three different objective measures of performance.