Paper Title
Evaluating the Credit Risk of Bank Customers by Combined Method of Support Vector Machine and Algorithm of Particles Swarm Optimization

Abstract
In the problems of risk identification, the imbalance of bank customer’s information should be considered and a remedy should be adopted for them; because, it is very effective on the model efficiency. Also, the known algorithms with supervisor like neural networks were used for improvement of results. The time of learning and constructing the model in the neural networks is longer, but, the time of testing it is short. Imbalance of data is very effective on learning in neural networks and accuracy of its results, and the rate of learning the data sample with less size is lower. In this research, a combinative method including support vector machine of particles swarm optimization was presented so that through it, we can discern more number of risks with more accuracy and less cost. The results of this research show that by using this combination, the accuracy of this method has been increased more than the neural network method and basic support vector machine and also, the discernment rate has been optimized in comparison with two main methods. Keywords - Particles Swarm Optimization, Credit Risk, Customers and Support Vector Machine