Hadoop Based Parallel Binary Bat Algorithm for Network Intrusion Detection

In Internet applications, due to the growth of big data with more features, intrusion detection has become a difficult process in terms of computational complexity, storage efficiency and getting optimized solutions of classification through existing sequential computing environment. Using a paralle...

Full description

Saved in:
Bibliographic Details
Published in:International journal of parallel programming Vol. 45; no. 5; pp. 1194 - 1213
Main Authors: Natesan, P., Rajalaxmi, R. R., Gowrison, G., Balasubramanie, P.
Format: Journal Article
Language:English
Published: New York Springer US 01.10.2017
Springer Nature B.V
Subjects:
Algorithms
Bayesian analysis
Big Data
Classification
Complexity
Computational efficiency
Computer networks
Computer Science
Computing time
Cybersecurity
Data management
Feature selection
Intrusion detection systems
Parallel processing
Processor Architectures
Software Engineering/Programming and Operating Systems
Studies
Theory of Computation
Feature selection
Parallel Binary Bat
Hadoop
Classification
MapReduce
ISSN:0885-7458, 1573-7640
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In Internet applications, due to the growth of big data with more features, intrusion detection has become a difficult process in terms of computational complexity, storage efficiency and getting optimized solutions of classification through existing sequential computing environment. Using a parallel computing model and a nature inspired feature selection technique, a Hadoop Based Parallel Binary Bat Algorithm method is proposed for efficient feature selection and classification in order to obtain optimized detection rate. The MapReduce programming model of Hadoop improves computational complexity, the Parallel Binary Bat algorithm optimizes the prominent features selection and parallel Naïve Bayes provide cost-effective classification. The experimental results show that the proposed methodologies perform competently better than sequential computing approaches on massive data and the computational complexity is significantly reduced for feature selection as well as classification in big data applications.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0885-7458
1573-7640
DOI:10.1007/s10766-016-0456-z