查看文章

Since the past decades, internet usage has become inevitable due to its tremendous applications in various fields. Due to this huge usage of network, a lot of security problems arise. Intrusion detection system (IDS) monitors the network events and filters the abnormal activities. While monitoring events, large amount of data samples are collected from sensors and the features are extracted from raw data which are required for IDS classification. This selection of best features from the raw data can be performed by the optimal feature selection method. To compute the detection accuracy, SVM classifier is used. The proposed model is tested using KDD99 benchmark dataset. Compared to other machine learning algorithm, the proposed method produced better results.