Enhanced Tree Based Real Time Intrusion Detection System in Big Data

S.J.SATHISH AARON JOSEPH; R. BALASUBRAMANIAN

doi:10.24297/ijct.v15i3.1671

Authors

S.J.SATHISH AARON JOSEPH RESEARCH SCHOLAR / HEAD , J.J.COLLEGE OF ARTS AND SCIENCE (AUTONOMOUS),PUDUKKOTTAI,TAMIL NADU
R. BALASUBRAMANIAN PROFESSOR , P.G AND RESEARCH J.J.COLLEGE OF ARTS AND SCIENCE (AUTONOMOUS) PUDUKKOTTAI , TAMIL NADU

DOI:

https://doi.org/10.24297/ijct.v15i3.1671

Keywords:

Classification, Data Imbalance, Decision Trees, Intrusion Detection, Random Forest

Abstract

Intrusion detection is one of the major necessities of the current networked environment, where every information is available in its corresponding digital form. This paper presents an enhanced tree based approach that can be used to perform intrusion detection faster and with better accuracy. The training data is subject to the random forest algorithm. This algorithm is a combination of tree predictors, and each tree depends upon the random vector generated. Spark based implementations of the Random Forest algorithm is used in a Hadoop cluster on datasets with varied imbalance to obtain the results. It has been observed that the classifier provided results in real time with an accuracy >90%, hence is more appropriate for online intrusion detection.