A Hybrid Model for Intrusion Detection Based on Genetic Clustering and PSO Algorithm

ABSTRACT:-  Reduction and selection of intruder attribute in intrusion detection system play an important role in process of detection. The huge number of attribute in intruder induces a problem in detection process and increase more time in decision making process. In this paper we tried to propose a very simple and fast clustering method for intrusion detection. A hybrid scheme based on coupling two different algorithms one is particle of swarm optimization and other is k-means algorithm. The main originality of proposed approach relies on associating two techniques: extracting more information bits via specific linguistic techniques, space reduction mechanisms, and moreover arcing cluster to aggregate the best clustering result. For the validation and performance evaluation of proposed algorithm used MATLAB software and KDDCUP99 dataset 10%. This dataset contains approx 5 lacks number of instance. The process of result shows that better detection ratio in compare of k-means and k-means-GA technique of intrusion detection.

Keyword: - Feature selection, Intrusion detection system, Genetic Algorithm, Clustering.

Publication Details - 

Title	A Hybrid Model for Intrusion Detection Based on Genetic Clustering and PSO Algorithm
Co-Author	Gaurav Shrivastava**
Publications	International journal of Master of Engineering Research and Technology
Date & Year	SEP 2015
Volume	Vol-2, Issue-9, 2015 Page No. 155-160
ISSN No.	ISSN 2394-6172

Download Full Length Paper

A review on Attribute Selection for Intrusion Detection System with Evolutionary Algorithm

ABSTRACT:- Reduction and selection of intruder attribute in intrusion detection system play an important role in process of detection. The huge number of attribute in intruder induces a problem in detection process and increase more time in decision making process. In this paper we tried to propose a very simple and fast clustering method for intrusion detection. A hybrid scheme based on coupling two different algorithms one is particle of swarm optimization and other is k-means algorithm. The main originality of proposed approach relies on associating two techniques: extracting more information bits via specific linguistic techniques, space reduction mechanisms, and moreover arcing cluster to aggregate the best clustering result. For the validation and performance evaluation of proposed algorithm used MATLAB software and KDDCUP99 dataset 10%. This dataset contains approx 5 lacks number of instance. The process of result shows that better detection ratio in compare of k-means and k-means-GA technique of intrusion detection. 

Keyword: - Feature selection, Intrusion detection system, Genetic Algorithm, Clustering.

Title	A review on Attribute Selection for Intrusion Detection System with Evolutionary Algorithm
Co-Author	Gaurav Shrivastava**
Publications	Engineering Universe for Scientific Research and Management
Date & Year	August 2015
Volume	Vol-7, Issue-8, 2015 Page No. 1-5
ISSN No.	ISSN 2394-6172

Download Full Lenght Paper

A Novel Method of Video Steganography Using Variable LSB

Abstract -

Abstract--Information hiding has become the focus of research today. The term Steganography means “covered writing” and involves transmission of secret messages through apparently innocent files without detection of the fact that message was sent. The innocuous file is known as the cover (video) while the file containing the hidden message referred as Stego medium. Video Steganography is the method of covert some secret data inside a video. In this paper, a novel approach of video Steganography is given which emphasis hiding secure data in video frames.

Keywords - Steganography, covert communication, ANN, FCNN

Publication Details - 

Title	A Novel Method of Video Steganography Using Variable LSB
Co-Author	Gaurav Shrivastava**
Publications	International Journal of Emerging Technology and Advanced Engineering
Date & Year	July 2015
Volume	Volume 5, Issue 7,  Page No. 462-464
ISSN No.	ISSN 2250-2459

Download Full Length Paper