A Novel Feature Selection Method Based on Trochoid Search Optimization with an S-Z-Shaped Composite X-Shaped Transfer Function for for DDoS Attack Detection.

Saved in:

Bibliographic Details
Title:	A Novel Feature Selection Method Based on Trochoid Search Optimization with an S-Z-Shaped Composite X-Shaped Transfer Function for for DDoS Attack Detection.
Authors:	Jin, Si-Yu¹ jsy@stu.ustl.edu.cn, Xing, Cheng² xingcheng0811@163.com, Wang, Jie-Sheng³ wjs@ustl.edu.cn, Xu, Zi-Rui¹ xzr@stu.ustl.edu.cn, Si, Zhi-Guang¹ szg@stu.ustl.edu.cn
Source:	Engineering Letters. Jul2026, Vol. 34 Issue 7, p2498-2515. 18p.
Subjects:	Feature selection, Denial of service attacks, Transfer functions, Optimization algorithms, Combinatorial optimization, Detection algorithms, Computer network traffic
Abstract:	Feature selection is a crucial step in classification tasks, capable of filtering out a subset of features that are more discriminative than the original features, thereby enhancing the model's generalization ability and reducing computational complexity. In DDoS attack detection, network traffic data typically exhibits characteristics of high dimensionality, strong redundancy and abundant noise, which directly impact detection efficiency and recognition accuracy. Therefore, efficient and stable feature selection is particularly important. Since the feature subset can be represented as a 0 and 1 vector, the feature selection problem is usually a binary combinatorial optimization problem. For this reason, a trochoid search optimization algorithm based on a composite X-shaped transfer function is proposed. This method, based on the optimization update mechanism of trochoid search, constructs a composite X-shaped transfer function that integrates the boundary suppression ability of the Z-shaped function and the smooth mapping characteristics of the S-shaped function, achieving a stable mapping from continuous solutions to binary feature subsets, thereby enhancing the balance between global exploration and local exploitation, reducing the risk of premature convergence, and improving the stability and repeatability of binary decisions. Through comparative experiments on 9 UCI datasets and 2 DDoS attack datasets, the effectiveness and applicability of this method have been verified. [ABSTRACT FROM AUTHOR]
	Copyright of Engineering Letters is the property of International Association of Engineers (IAENG) and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	Feature selection is a crucial step in classification tasks, capable of filtering out a subset of features that are more discriminative than the original features, thereby enhancing the model's generalization ability and reducing computational complexity. In DDoS attack detection, network traffic data typically exhibits characteristics of high dimensionality, strong redundancy and abundant noise, which directly impact detection efficiency and recognition accuracy. Therefore, efficient and stable feature selection is particularly important. Since the feature subset can be represented as a 0 and 1 vector, the feature selection problem is usually a binary combinatorial optimization problem. For this reason, a trochoid search optimization algorithm based on a composite X-shaped transfer function is proposed. This method, based on the optimization update mechanism of trochoid search, constructs a composite X-shaped transfer function that integrates the boundary suppression ability of the Z-shaped function and the smooth mapping characteristics of the S-shaped function, achieving a stable mapping from continuous solutions to binary feature subsets, thereby enhancing the balance between global exploration and local exploitation, reducing the risk of premature convergence, and improving the stability and repeatability of binary decisions. Through comparative experiments on 9 UCI datasets and 2 DDoS attack datasets, the effectiveness and applicability of this method have been verified. [ABSTRACT FROM AUTHOR]
ISSN:	1816093X