A Novel Mixture Classifier and its Application in Breast Cancer Prognosis

In clinical applications, such as breast cancer prognosis, many classical factors are generally considered not to be sufficient for our need. Along with micro-arrays have had a great impact on cancer research in the past decade, new and more powerful gene biomarkers are discovered. And advanced prognostic index are still called for yet. In another way, currently many classification method/strategy, which could be used in above purpose, would be wasteful to lose much invaluable information in abnegated data when they pursuit for obtaining higher predict accuracy. So this paper designs and realizes a novel mixture classification model by combining the developing rough set classification and currently strongest SVM classification. This new classifier is used as a combined prognostic method for breast cancer prognosis and gets better effect than previous research.