Denial of Services Attack Detection using Random Forest Classifier with Information Gain

  Reena Singh Rajput,  Dr. Sanjay Agrawal

Abstract- Denial of service attacks (DoS) is a common threat to many online services. These attacks aim to overcome the availability of an online service with massive traffic from multiple sources. Denial of Service (DoS) is a prevalent threat in today’s networks. Denial of service attacks is the very common problem in the present scenario. To get rid of DoS attack we have the intrusion detection systems but we need to maintain the performance of the intrusion detection systems. Therefore, we propose a novel model for intrusion detection system using random forest classifier and Information Gain (IG) model. Random Forest (RF) is an ensemble classifier and performs well compared to other traditional classifiers for effective classification of attacks. Intrusion detection system is made fast and efficient by use of optimal feature subset selection using IG. In this model we have tried to find out an optimal feature subset that gives performance greater than or equal to the performance given by the set of 41 features and time taken to build the model by the selected feature set is less than the time taken by the set of 41 features. This makes the intrusion detection systems faster and efficient. To evaluate the performance of our model, we conducted experiments on NSL-KDD data set. Empirical result shows that proposed model is efficient, fast and robust and can get the high accuracy in detection DoS attack using WEKA tool.

Keywords- Denial of Service (DoS), NSL-KDD dataset, Random Forest, Information Gain.

Unique Identification Number - IJEDR1703132
Page Number(s) - 929-938
Pubished in - Volume 5 | Issue 3 | September 2017
DOI (Digital Object Identifier) -   
Publisher - IJEDR (ISSN - 2321-9939)

  Reena Singh Rajput,  Dr. Sanjay Agrawal,   "Denial of Services Attack Detection using Random Forest Classifier with Information Gain", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.5, Issue 3, pp.929-938, September 2017, Available at :http://www.ijedr.org/papers/IJEDR1703132.pdf