Optimizing the Classification Cost using SVMs with a Double Hinge Loss

The objective of this study is to minimize the classification cost using Support Vector Machines (SVMs) Classifier with a double hinge loss. Such binary classifiers have the option to reject observations when the cost of rejection is lower than that of misclassification. To train this classifier, the standard SVM optimization problem was modified by minimizing a double hinge loss function considered as a surrogate convex loss function. The impact of such classifier is illustrated on several discussed results obtained with artificial data and medical data. Povzetek: Predstavljena je optimizacija cene klasificiranja z metodo strojnega učenja SVM.