An Exploratory Analysis of Feature Selection for Malware Detection with Simple Machine Learning Algorithms - Original scientific article

Aug 4, 2026

Upload by Finn

Computers face rising risk from malware and other hostile attacks, while current malware detection techniques often struggle as analytical data grows and many irrelevant attributes reduce detection effectiveness. This study proposes a feature selection approach to improve malware detection accuracy by identifying pertinent and significant traits, while reducing computational cost. Most informative features are derived from multiple machine learning methods, followed by six classifiers evaluated on two malware datasets. Results indicate RF and DT outperform other techniques in accuracy, precision, F1-score, and recall.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Abstract

Index Terms

I. Introduction

An Exploratory Analysis of Feature Selection for Malware Detection with Simple Machine Learning Algorithms Md Ashikur Rahman, Syful Islam, Yusuf Sulistyo Nugroho, Member, IEEE, Fatah Yasin Al Irsyadi, and Md Javed Hossain* Original scientific article Abstract—Computers have become increasingly vulnerable to malicious attacks with an increase in popularity and the proliferation of open system architectures. There are numerous malware detection technologies available to protect the computer operating system from such attacks. This type of malware detector targets programs based on patterns detected in the properties of computer applications. As the amount of analytical data increases, the computer defense system is adversely affected. The performance of the detection mechanism has been hindered due to the presence of numerous irrelevant characteristics. The goal of this study is to provide a feature selection approach that will help malware detection systems be more accurate by detecting pertinent and significant traits. Furthermore, by selecting the most important features, it is possible to maintain an acceptable level of accuracy in the detection of malware while significantly lowering the computational cost. The proposed method displays the most important features (MIFs) obtained from each machine learning method, including data cleaning and feature selection. Furthermore, the method applies six machine learning classification techniques to the selected feature set. Several classifiers were evaluated based on several characteristics for malware detection, including Support Vector Machines (SVM), Logistic Regression (LR), K-nearest neighbor (K-NN), Decision Tree (DT), Naive Bayes (NB), and Random Forest (RF). Our suggested model was tested on two malware datasets to determine its effectiveness. In terms of accuracy, precision, F1 scores, and recall, the experimental findings show that RF and DT classifiers beat other techniques. Index Terms—Malware Detection, Machine Learning, Feature Selection, Information Gain, Cybersecurity. I. INTRODUCTION Malware stands for malicious software and is intended to harm systems and users by stealing information, destroying data, or simply irritating them. Malware is widely distributed and computer security issues are on the rise, according to reports [1] . Malicious programs or attacks, such as malware and ransomware families, continue to pose serious cybersecurity Manuscript received June 20, 2023; revised June 30, 2023 . Date of publication September 13, 2023 . Date of current version September 13, 2023 . M. A. Rahman and M. J. Hossain are with the Noakhali Science and Technology University, Bangladesh ([mdashikur567@gmail.com](mdashikur567@gmail.com), [javed@nstu.edu.bd](javed@nstu.edu.bd)). S. Islam is with the Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Bangladesh ([syfulcse@bsmrstu.edu.bd](syfulcse@bsmrstu.edu.bd)). Y. S. Nugroho and F. Y. Al Irsyadi are with the Universitas Muhammadiyah Surakarta, Indonesia ({yusuf.nugroho, [fatah.yasin](fatah.yasin}@ums.ac.id)[}](fatah.yasin}@ums.ac.id)[@ums.ac.id](fatah.yasin}@ums.ac.id)). Digital Object Identifier (DOI): 10.24138/jcomss-2023-0091 *Corresponding author concerns, with potentially devastating effects [2] . Computer systems, data centers, the Web, and mobile devices have been damaged, with applications in a variety of sectors and companies [3], [2], [4] . By encrypting data in an unbreakable format that can only be decoded by the attacker, the majority of malware is designed to restrict victims from accessing computer data [5] . Since removing the infection results in irreversible damage, victims are compelled to comply with the attacker’s demands [6] . If anyone does not comply with the attacker’s demands, data will be permanently gone. Assailants are using current technology to turn traditional malware into developing malware families, making it more difficult to recover from a m

FAQs

Why is feature selection important for malware detection in this study?

As analytical data increases, detection performance degrades due to many irrelevant characteristics. Feature selection helps focus on significant traits to improve accuracy and lower computational cost.

What machine learning classifiers were evaluated?

The study evaluates Support Vector Machines (SVM), Logistic Regression (LR), K-nearest neighbor (K-NN), Decision Tree (DT), Naive Bayes (NB), and Random Forest (RF).

How does the proposed workflow support malware detection?

The method cleans data, performs feature selection to obtain the most important features, and then applies multiple classifiers to the selected feature set. It is tested on two malware datasets to measure effectiveness.

Aug 4, 2026

Upload by Finn