![]() | Only 14 pages are availabe for public view |
Abstract According to rapid development and popularity of Internet and online procedures, the potential of network attacks has increased substantially in recent years. Therefore, network security needs to be concerned to provide secure information channels. Intrusion Detection System (IDS) becomes an essential component of computer security . Network Intrusion Detection Systems (NIDS) aims to dynamically identify unusual access or attacks to secure the internal networks, by looking for potential malicious activities in network trac. However, building a high-performance and fast NIDS is a major research problem in network security. One of the important problems for NIDS is dealing with data containing high number of features. High dimensional data may leads to decrease the predictive accuracy and the speed of the NIDS. Therefore, Feature Selection (FS) is one of the key topics in building NIDS. (FS) can serve as a preprocessing tool for high dimensional data before solving the classication problems. The purpose of the feature selection is to reduce the number of irrelevant and redundant features. (FS) searches for a subset of features which improve the prediction accuracy and improves the NIDS speed. This thesis is devoted to focuss on how to construct a fast accurate NIDS. The thesis propose two dierent hybrid NIDS, the proposed hybrid NIDS models involves data preprocessing, data reduction and intrusion classification. Experiments and Analysis of the proposed hybrid NIDSs with other previous NIDSs demonstrated that; the two proposed hybrid NIDSs enhance the intrusion detection rate and decreasing the testing speed. |