Improving Support Vector Machine For Learning Unbalanced Data Set

Support Vector Machines (SVM) was proposed by Vapnik in1990'. SVM is a new and outstanding learning machine and is an efficient machine-learning tool in dealing with small samples. SVM has been widely applied to many areas, such as pattern recognition,signal Processing, automation, communication,etc.In the unbalanced sets, the difference of the sample quantities of the different classes leads to declining performances of many classifiers. So the unbalanced sets is always the research hotspot in machine learning. Seeking for the optimal hyperparameters with the unbalanced sets is one of the most important branches of SVM and often named as model selection.In practice, training data is usually unbalanced, one class is"rare"relative to the other, and misclassification cost of the rare class may be much greater than the cost of the other class. In this situation, accuracy and the misclassification cost should be considered. This paper mainly studies the optimal methods for parameter selection in SVM for unbalance data sets.SVM has been used in various fields and has obtained good effects; the parameter selection in SVM is an important research direction, different parameters result in different generalization; The studies of parameter selection in SVM for unbalance data sets is fewer. For unbalance data sets, this paper presented the parameter selection model in SVM and algorithm, done the experiments.In this paper, the SVM learning method is extended, based on the Gauss kernel, by use of C+( the weight assigned to the rare class), and C- (the weight assigned to the other class)to train more sensitive hyperplane, which is optimized by generic algorithm. Meanwhile, a new sensitive quality measure function is introduced in the optimization process. The experiment results show that the optimized algorithm has competitive performance when dealing with the rare class in the unbalance training data.