Journal of Computer & Information Technology

Anomaly Detection Using Artificial Neural Network

Author:
MANORANJAN PRADHAN1, SATEESH KUMAR PRADHAN2 and SUDHIR KUMAR SAHU3
Affiliation:

1Department of Computer Science & Engg., GITA, Bhubaneswar, Odisha (INDIA)
2Department of Computer Science, Utkal University, Odisha (INDIA)
3P.G. Department of Statistics, Sambalpur University, Odisha (INDIA)

Keyword:
neural network
Issue Date:
December 2012
Abstract:
In this research, anomaly detection using neural network is introduced. This research aims to experiment with user behaviour as parameters in anomaly intrusion detection using a backpropagation neural network. Here we wanted to see if a neural network is able to classify normal traffic correctly, and detect known and unknown attacks without using a huge amount of training data. For the training and testing of the neural network, we used the DARPA Intrusion Detection Evaluation data sets. In our final experiment, we have got a classification rate of 88% on known and unknown attacks. Compared with other researches our result is very promising.
Pages:
ISSN:
2455-9997 (Online) - 2229-3531 (Print)
Source:
DOI:

Copy the following to cite this article:

Ansari Education And Research Society
Facebook Google Plus Twitter