A boosting method for maximization of the area under the ROC curve

摘要: We discuss receiver operating characteristic (ROC) curve and the area under ROC (AUC) for binary classification problems in clinical fields. propose a statistical method combining multiple feature variables, based on boosting algorithm maximization of AUC. In this iterative procedure, various simple classifiers that consist variables are combined flexibly into single strong classifier. consider regularization to prevent overfitting data using penalty term nonsmoothness. This not only improves performance but also helps us get clearer understanding about how each variable is related outcome variable. demonstrate usefulness score plots constructed componentwise by method. describe two simulation studies real analysis order illustrate utility our