A BI-OBJECTIVE HYPER-HEURISTIC SUPPORT VECTOR MACHINES FOR BIG DATA CYBER-SECURITY

Authors

  • B. VENKATESH Author
  • PATHIKONDA SATWIKA REDDY Author
  • GANDLA SAIKUMAR Author
  • PAMARTHI ASHOK KUMAR Author
  • VEMULAPALLI HARISCHANDRA PRASADU Author

Keywords:

The SVM configuration search, space is efficiently explored, low-level  heuristics

Abstract

Researchers have a formidable task in 
addressing cyber security concerns in the context 
of big data. It has been proposed that machine 
learning techniques may be used to deal with 
large data security issues. Support vector 
machines (SVMs) are one such approach that has 
shown exceptional performance on several 
classification issues. However, in order to set up 
an efficient SVM, the user must first determine 
the optimal SVM configuration; this is a difficult 
operation that calls for specialised expertise and 
much trial and error. In this research, we provide 
a formalisation of the SVM configuration 
procedure as a bi-objective optimisation problem, 
where accuracy and model complexity are treated 
as competing goals. We present a new problemdomain-agnostic hyper-heuristic framework for 
bi-objective optimisation. For the first time, a 
hyper-heuristic has been created specifically for 
this issue. High-level strategy and low-level
heuristics are what make up the suggested hyperheuristic framework. In order to determine which 
of many possible low-level heuristics should be 
utilised to produce a new SVM configuration, the 
high-level approach monitors search 
performance. The SVM configuration search 
space is efficiently explored by the low-level 
heuristics, each of which follows a unique set of 
rules. The proposed framework adaptively
merges the benefits of decomposition- and 
Pareto-based techniques to approximate the 
Pareto set of SVM configurations, making it 
suitable for bi-objective optimisation. The 
suggested methodology has been tested on two 
cyber security issues: classifying Microsoft 
malware using a large data set and detecting 
anomalies in network traffic. In comparison to its 
contemporaries and other algorithms, the
acquired results show that the suggested 
framework is very successful

Downloads

Published

01-06-2024

How to Cite

A BI-OBJECTIVE HYPER-HEURISTIC SUPPORT VECTOR MACHINES FOR BIG DATA CYBER-SECURITY. (2024). International Journal of Engineering Research and Science & Technology, 20(2), 489-498. https://ijerst.org/index.php/ijerst/article/view/312