Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
3 results
Search Results
Item Restricted Insider Threat Detection in a Hybrid IT Environment Using Unsupervised Anomaly Detection Techniques(Saudi Digital Library, 2025) Alharbi, Mohammed; Antonio, GouglidisThis dissertation analyses insider threat detection in hybrid IT environments with unsupervised anomaly detection techniques. Insider threats, including those committed by trusted persons with granted access, are considered to be one of the most challenging to alleviate cybersecurity threats because they resemble legal user behavior and do not have labelled datasets to train supervised models. Hybrid infrastructures, an integration of on-premise and cloud resources, also make detection harder as they create large, heterogeneous and fragmented logs. In order to cope with such challenges, this paper presents a detection system that uses isolation forest and local outlier factor algorithms. Multi-source organisational data, such as authentication, file, email, HTTP, device and LDAP logs, were pre-processed and loaded into enriched user profiles, with psychometric attributes added where possible. The framework was assessed by the CERT Insider Threat Dataset v6.2, where the results indicated that both algorithms were effective in detecting anomalous behaviours: Isolation Forest was effective in detecting global outliers, whereas Local Outlier Factor was good in detecting subtle local outliers. It was found through the comparative analysis that the strength of each method was complementary, and they should be used together when stratifying users into high-, medium-, and low-risk groups. Although it still has constraints in terms of synthetic data, real-time implementation, and ecological validity, the study is relevant in the development of anomaly-based detection methods and offers viable information to organisations wishing to be proactive in curbing insider threats61 0Item Restricted Harnessing Machine Learning and Deep Learning for Analyzing Electrical Load Patterns to Identify Energy Loss(Saudi Digital Library, 2025) Alabbas, Mashhour Sadun Abdulkarim; Albatah, MohammadMeeting the challenges of energy requirements, consumption patterns, and the push for sustainability makes energy management in contemporary agriculture critically important. This study aims to devise a holistic model for energy efficiency in agricultural contexts by integrating modern computer vision methodologies for field boundary extraction together with anomaly detection techniques. To achieve the accurate segmentation of agricultural fields from satellite imagery, high-resolution imagery is processed using the YOLOv8 object detection model. The subsequently generated field feature datasets enable the smart grid data to serve as a basis for the anomaly detection process using the Isolation Forest algorithm. The methodology follows a multi-stage pipeline: data collection, preprocessing, augmentation, model training, fine-tuning, and evaluation. To validate accurate and reliable field boundary detection, evaluation metrics precision, recall, and mAP (mean Average Precision) are computed and analyzed. Subsequently, energy consumption data are processed for anomaly detection, enabling the identification of irregular and potentially inefficient consumption patterns. The findings indicate that YOLOv8 has a very high detection accuracy with an mAP score over 90%. Furthermore, the Isolation Forest algorithm has shown improved F1 scores over traditional approaches in detecting anomalies in energy consumption. This integrated method provides an automated and scalable solution in precision agriculture which allows users to monitor cultivation conditions and minimize energy consumption, thereby enhancing the energy efficiency and the overall decision-making framework. The study advances the convergence of artificial intelligence, remote sensing, and intelligent energy management systems, offering a basis for developing technological innovations that promote sustainablility in agriculture.44 0Item Restricted Credit Card Fraud Prediction Using Machine Learning Model(University of Essex, 2024-08) Alanazi, Mohammed; Walton, MichaelThe widespread adoption of credit cards has significantly increased the frequency of fraudulent activities. This has resulted in considerable financial losses for both consumers and financial institutions. As the use of credit cards continues to grow, the challenge of protecting transactions against unauthorized access has become more serious than ever. This research focuses on creating a solution using machine learning to accurately and effectively identify fraudulent credit card transactions. It addresses the issue of uneven transaction data by employing advanced methods such as logistic regression, XGBoost, LightGBM, and a hybrid model. The research involves thorough data preparation, model development, and careful assessment using measures “such as accuracy, precision, recall, F1 score, and ROC AUC”. This research leverages sophisticated machine learning techniques and tackles the specific challenges associated with imbalanced data. The study aims to significantly enhance the detection of fraudulent transactions while reducing false positives. The ultimate goal is to boost the security of financial systems, thus providing better protection against fraud, and to improve trust and reliability in credit card transactions.65 0