A hybridization of butterfly optimization algorithm and harmony search for fuzzy modelling in phishing attack detection

Fuzzy system is one of the most used systems in the decision-making and classification method as it is easy to understand because the way this system works is closer to how humans think. It is a system that uses human experts to hold the membership values to make decisions. However, it is hard to de...

Full description

Saved in:
Bibliographic Details
Published in:Neural computing & applications 2023-03, Vol.35 (7), p.5501-5512
Main Authors: Nordin, Noor Syahirah, Ismail, Mohd Arfian
Format: Article
Language:English
Subjects:
Algorithms
Artificial Intelligence
Computational Biology/Bioinformatics
Computational Science and Engineering
Computer Science
Data Mining and Knowledge Discovery
Datasets
Decision making
Image Processing and Computer Vision
Machine learning
Mathematical models
Modelling
Optimization
Optimization algorithms
Original Article
Phishing
Probability and Statistics in Computer Science
Process parameters
Websites
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Fuzzy system is one of the most used systems in the decision-making and classification method as it is easy to understand because the way this system works is closer to how humans think. It is a system that uses human experts to hold the membership values to make decisions. However, it is hard to determine the fuzzy parameter manually in a complex problem, and the process of generating the parameter is called fuzzy modelling. Therefore, an optimization method is needed to solve this issue, and one of the best methods to be applied is Butterfly Optimization Algorithm. In this paper, BOA was improvised by combining this algorithm with Harmony Search (HS) in order to achieve optimal results in fuzzy modelling. The advantages of both algorithms are used to balance the exploration and exploitation in the searching process. Two datasets from UCI machine learning were used: Website Phishing Dataset and Phishing Websites Dataset. As a result, the average accuracy for WPD and PWD was 98.69% and 98.80%, respectively. In conclusion, the proposed method shows promising and effective results compared to other methods.
ISSN:0941-0643
1433-3058
DOI:10.1007/s00521-022-07957-0