Adaptive Asymmetrical SVM and Genetic Algorithms Based Iris Recognition

We propose Genetic Algorithms to improve the feature subset selection by combining the valuable outcomes from multiple feature selection methods. This paper also motivates the use of asymmetrical SVM, which focuses on two important issues. The first issue is the sample ratio bias, and the second issue is that the different types of misclassification error may have different costs, which lead to different misclassification losses. The asymmetrical SVM also influences the trade-off between the cases of False Accept and False Reject. In order to overcome the problem induced by the traditional SVM due to its slower performance issue in the test phase caused by the number of support vectors, we also implement an adaptive algorithm to select the feature vector (FV) from the support vector solutions.