A Rough Set based Feature Selection Algorithm for Effective Intrusion Detection in Cloud Model

There exist many problems in intrusion detection systems such as large data volume, features and data redundancy which seriously affect the efficiency of the detection algorithm. Such problems need to be addressed in developing reliable intrusion detection systems. In this paper, we propose an intrusion detection model that combines Rough Set based Feature Selection Algorithm and Fuzzy SVM for effective intrusion detection in the Cloud. The algorithm evaluates the characteristics of the attribute weights for security in the Cloud model and generates an optimal number of features in order to achieve best trade-off between detection rate and false alarm rate by using rough sets. This model solves the problem of feature redundancy and hence helps to improve the speed of evaluation. The experimental results obtained from this work show that the proposed model solves the feature selection problem and classification for intrusion detection effectively. It also achieves better detection performance for various types of attacks with reduced detection time.