MACHINE LEARNING IN CYBERSECURITY - EXPLORING APPROACHES FOR MALWARE DETECTION AND CATEGORIZATION - Master’s Degree Thesis

Sep 17, 2026

Uploaded by Ophelia

This study focuses on the complex task of malware detection by creating, assessing, and contrasting machine learning and deep learning models to identify and categorize malware in both text files and executable (.EXE) file types. It uses a quantitative methodology with two main datasets: Dike for portable executable and object linking and embedding files, and a malware categorization dataset for .txt formats. Models including ANN, Random Forest, AdaBoost, and CNN1D are evaluated using accuracy, precision, recall, F1 score, and confusion matrices.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages54

Reading Time2 h 16 min

AccessFree to download

Abstract

1. Introduction

1.1. Background

1.2. Problem Statement

1.3. Objectives of the Study

1.4. Thesis Structure

2. Review of the Litrature

2.1. A Historical Viewpoint on Malware Detection

2.2. Significance of Machine Learning in Malware Detection and Cybersecurity

2.3. Correlated Works

2.4. Contribution of this Study

3. Methodology

3.1. Overview of the Approach

3.2. Data Collection

3.3. Machine Learning Methods

3.4. Deep Learning Methods

3.5. Evaluation Metrics

T.R. USKUDAR UNIVERSITY INSTITUTE OF SCIENCE DEPARTMENT OF GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES MASTER'S DEGREE PROGRAM OF CYBER SECURITY MASTER'S DEGREE THESIS MACHINE LEARNING IN CYBERSECURITY: EXPLORING APPROACHES FOR MALWARE DETECTION AND CATEGORIZATION Saqib Shabir Peerzada Advisor Asst.Prof.Dr.Ahmet ŞENOL ISTANBUL-2024 T.R. USKUDAR UNIVERSITY INSTITUTE OF SCIENCE DEPARTMENT OF GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES MASTER'S DEGREE PROGRAM OF CYBER SECURITY MASTER'S DEGREE THESIS MACHINE LEARNING IN CYBERSECURITY: EXPLORING APPROACHES FOR MALWARE DETECTION AND CATEGORIZATION Saqib Shabir Peerzada Advisor Asst.Prof.Dr.Ahmet ŞENOL ISTANBUL-2024 ABSTRACT This study focuses on the complex task of malware detection by creating, assessing, and contrasting machine learning and deep learning models for the identification and categorization of malware contained in both text files as well as executable (.EXE) file types. The research employs a rigorous quantitative research methodology and utilises two main datasets. One of them being Dike Dataset for Portable executable and object linking and embedding files, and the Malware Categorization Dataset for .txt file formats. This study will utilize Neural Networks (ANN), Artificial Neural Networks (ANN), Random Forest Classifier, AdaBoost Classifier, and Convolutional Neural Networks (CNN1D), are thoroughly evaluated and analysed using metrics such as accuracy, precision, recall, F1 score, and confusion matrices. The study demonstrates the improved effectiveness of the HistGradientBoosting Classifier and Random Forest Classifier in analysing text and EXE files, respectively. It also compares their performance with that of ANN and CNN1D models. The integration of these models into a user-centric Streamlit application expands the availability of sophisticated virus detection approaches. The results highlight the strong capacity of integrating machine learning and deep learning to enhance cybersecurity technologies and processes. This incorporation into a functional application represents a significant improvement in cybersecurity measures designed for both personal and institutional utilisation. Research Keywords: Malware Detection, Machine Learning, Deep Learning, Cybersecurity, Comparative Analysis, Artificial Neural Networks, Convolutional Neural Networks, Streamlit TABLE OF CONTENTS Abstract .................................................................................................................... i 1. Introduction .................................................................................................... 1 1.1. Background ............................................................................................. 1 1.1.1. Importance of Malware Detection ...................................................... 1 1.1.2. Evolution of Malware and Detection Techniques .............................. 1 1.2. Problem Statement .................................................................................. 1 1.2.1. Challenges in Current Malware Detection Methods .......................... 1 1.2.2. The Imperative for Employing ML in Malware Recognition ............ 2 1.3. Objectives of the Study ........................................................................... 3 1.3.1. To Build a Model That Can Detect Malware from Files.................... 3 1.3.2. Model That Can Group Malware into Particular Categories.............. 3 1.3.3. Testing Various Models to Find the most accurate One..................... 3 1.3.4. Comparative Analysis Between Machine Learning and Deep Learning Models ...................................................................................................................... 3 1.3.4. Developing a User-Friendly Application............................................ 3 1.3.5. Contributing to the domain of Threat Detection and Cyber Security.

FAQs

Which datasets are used for malware detection and categorization in the study?

The study uses the Dike Dataset for portable executable and object linking and embedding files, and a Malware Categorization Dataset for .txt file formats.

What types of malware inputs are the models trained and evaluated on?

Models are designed to detect and categorize malware contained in both text files and executable (.EXE) file types.

Which evaluation metrics and model comparisons are reported?

Performance is assessed using accuracy, precision, recall, F1 score, and confusion matrices, comparing methods including ANN, Random Forest, AdaBoost, and CNN1D.

Sep 17, 2026

Uploaded by Ophelia

MACHINE LEARNING IN CYBERSECURITY - EXPLORING APPROACHES FOR MALWARE DETECTION AND CATEGORIZATION - Master’s Degree Thesis_Page 1

Open in WPS to view all 54 pages