Performance Analysis of Dimension Reduction Techniques With Classifier Combination for Intrusion Detection System

As increase in the internet services and usage with open access to sensitive data, necessity of security to these systems had become a need of the hour. Intrusion Detection Systems (IDSs) provide an important layer of security for computer systems and networks, and are becoming more and more crucial issue. To detect the attacks hitting the network it is very obligatory to properly monitor the flow of network data. Network data is pre-processed before classification. Feature Selection is one of the preprocessing technique which eliminates irrelevant features and select more informative features from the data set with respect to the task to be performed. Feature Selection techniques is adopted as it is computationally efficient, simplified, have more feature interpretability, less learning time, improves data relevancy. Feature Selection can be applied to both supervised and unsupervised learning methodologies. Feature Selection selects a subset of the features in the training set and further using this subset as features in classification. Feature extraction creates new features from functions of the original features, whereas feature selection returns a subset of the features from the dataset. On selecting the features, classification is performed for detecting the intrusion.