A This paper presents a novel one-class classification approach to intrusion detection based support vector data description. This approach is used to separate target class data from other possible outlier class data, which are unknown to us. SVDD-intrusion detection enables determination of an arbitrary shaped region that comprises a target class of a dataset. This paper analyzes the behavior of the classifier based on parameter selection and proposes a novel way based on genetic algorithm to determine the optimal parameters. Finally some experiments are finally reported with DARPA' 99 evaluation data. The results demonstrate that the proposed method outperforms other two-class classifiers.