The Solution to How to Select an Optimal Set of Features from Many Features Used to Intrusion Detection System in Wireless Sensor Network
Sheng Yu, Hui Nan Guo, Gui Xian Yu, Xu Jin, Li Nang Zhang, Tie Jun Shao · 2010
To improve the reality and whole performance of network Intrusion Detection System(IDS) ,the approach in which an optimal combination of ( > ) features used to classification of IDS were selected from, features, was presented, which was based on Genetic Algorithms. The optimal set of features was best for recognizing intruder, which made classification evaluation target to reach maximum. Tests also showed that the method and IDS (neural network) developed was useful and available. It is conclusion that to select an optimal set of features from many features and to reduce its number (the number should be proper) can lead to debasing computing complexity of IDS and times, so as to improve the whole performance of IDS.