BSpace

The British University in Dubai (BUiD) Digital Repository

Welcome to BSpace, the online institutional repository of the British University in Dubai. BSpace provides access to the Dissertations, Thesis, Research projects, Faculty publications and archives of BUiD.

Submit your dissertation/thesis by completing the registration using your BUiD email.

Review the submission guidelines before you submit the final version

Communities in DSpace

Select a community to browse its collections.

Now showing 1 - 7 of 7

Recent Submissions

  • Thumbnail Image
    Item type:Item,
    Leveraging Retrieval-Augmented Language Models for Early Diagnosis in Resource-Constrained Healthcare
    (The British University in Dubai (BUiD), 2025-01) ZOHAIR, LUBNA MAHMOUD ABU; Prof Abdallah, Sherief
    Large language models (LLMs) encounter notable challenges when applied to sensitive domains such as healthcare, particularly where data is limited, highly confidential, and subject to strict regulatory frameworks. These challenges are especially pronounced in the context of rare disease diagnosis, where current approaches often rely on decoder-based models that are proprietary and prone to hallucinations and generation of inaccurate or misleading outputs. Additionally, the substantial computational demands of LLMs further limit their feasibility in resource-constrained or low-income settings. To address these challenges, this research proposes a framework that maximizes the diagnostic utility of small and early collected clinical datasets while leveraging the power of open-source pre-trained medical language models. The framework introduces Retrieval-Augmented Encoding (RAE), a technique designed to enhance the diagnostic performance of affordable language models with classification heads by retrieving similar clinical notes to enrich the encoding of input data and support inference in diagnostic tasks. It also employs Retrieval-Augmented Generation (RAG) to expand the training dataset through paraphrasing for fine-tuning. A case study on appendicitis diagnosis was conducted using 2,400 unstructured abdominal disease notes, focusing on the exploration of the diagnostic sufficiency of early-stage clinical notes such as the History of Present Illness (HPI). Results show that the proposed framework achieved diagnostic accuracy and precision rates exceeding 93.3% with the HPI notes alone, highlighting their potential for early and efficient diagnosis without reliance on additional unstructured notes, like Physical Examination. This research highlights how locally deployable language models can be used effectively in resource-constrained healthcare environments to support early and accurate diagnoses, particularly for rare diseases and critical clinical decisions. Keywords: large language models, small dataset, appendicitis, diagnosis, retrieval augmented generation, data augmentation, clinical notes, history of present illnesses, BERT models, rare disease, healthcare informatics
  • Thumbnail Image
    Item type:Item,
    Cyber Threat Intelligence Framework for Enhancing the Robustness of AI Models Against Adversarial Machine Learning
    (The British University in Dubai (BUiD), 2025-09) ALTENEIJI, FAISAL; Prof Abdallah, Sherief
    Adversarial Machine Learning (AML) threats pose critical challenges to the robustness of Intrusion Detection System (IDS) models, as existing IDS often lack intelligence-driven mechanisms to anticipate and mitigate evolving adversarial machine learning attacks. This research addresses these challenges by developing a Cyber Threat Intelligence (CTI)-based framework to enhance the resilience of AI-driven IDS. The framework integrates activity attack graphs, similarity analysis across tactical, technical, operational, and strategic intelligence, and cross-intelligence comparison supported by historical threat data to identify evolving threats and control gaps. The study follows an explanatory mixed-methods design. In the quantitative phase, experimental adversarial attacks were conducted against three IDS models: Artificial Neural Networks (ANN), Support Vector Machine (SVM), and Decision Tree using the CICIDS2017 dataset. Findings demonstrated that certain activity threads (e.g., model poisoning and targeted evasion) shared overlapping indicators, adversarial operations, and control gaps across intelligence types, thereby validating the framework’s ability to uncover hidden relationships between adversarial threats. In the qualitative phase, semi-structured interviews with domain experts validated the framework, highlighting its strengths in integrating monitoring, analytics, and threat hunting while identifying executional challenges such as data quality, scalability, and ethical constraints. This research makes a novel contribution by integrating CTI practices with AML defence to improve IDS model’s robustness. The findings provide both theoretical insight and practical guidance for organizations seeking to implement intelligence-driven strategies to protect AI models against adversarial attacks. Keywords: adversarial machine learning, cyber threat intelligence, artificial intelligence, network security
  • Thumbnail Image
    Item type:Item,
    An Approximate Anytime Hierarchical Clustering Earth Mover’s Distance (AAHC-EMD)
    (The British University in Dubai (BUiD), 2025-06) ABDEL SALAM, MINAT ALLAH ESSAM HAMZA; Prof Sherief Abdallah
    Flow Cytometry (FC) is a crucial tool for analysing soluble substances such as blood, where added biomarkers help highlight any available abnormalities or diseases. This results in numerical datasets that are analysed to assess similarities or dissimilarities between samples. However, applying machine learning to hierarchical FC datasets presents challenges due to their two-level structure: the top level contains blood cells, while the bottom level holds cell attributes. In this study, the dataset was reduced from 30,000 cells to 2,500 and from 8 to 4 attributes per sample to manage time consumption. Despite this reduction, the analysis still required handling 10,000 attributes at the lower level, which is computationally impractical. While dimensionality reduction could help, it risks losing critical information. This thesis proposes treating each sample as a cluster configuration, using Earth Mover’s Distance (EMD), which is robust to instrumental drift but computationally expensive. The solution employs an Approximate Anytime Hierarchical Clustering-based EMD lower bound (AAHC-EMD) algorithm to calculate similarity by measuring the distance between cluster centroids instead of individual cells. This method ranks testing samples to each query at each stage of the hierarchy, reducing computational time by 48% to 89%, and achieves 100% ranking accuracy when given more time. Using the same approach also identifies the Best-Fit testing sample for each query from a list of 20 testing samples, producing 100% accuracy and a 72.5% time saving compared to traditional EMD. This approach improves diagnostic precision and computational efficiency in analysing complex hierarchical datasets such as FC. Keywords: Anytime Algorithm, Earth Mover’s Distance, Flow Cytometry, Hierarchical Clustering, k-means, Lower bound.
  • Thumbnail Image
    Item type:Item,
    Examining the Role of Innovative Leadership in Enhancing Student Achievement : strategies and Qutcomes
    (The British University in Dubai (BUiD), 2024-11) ALKENDI, MARWA; Dr Tendai, Charles
    This research study examines the influence of innovative leadership tactics on student performance, emphasizing approaches and outcomes within various educational settings. As educational environments become increasingly complex, school administrators need to address the issue of reconciling established traditional approaches with the necessity for innovative practices that improve learning outcomes. This study seeks to examine how innovative leadership strategies might foster a culture of continuing development, cooperation, and adaptation among educators, thereby augmenting student success. The study used a mixed-methods approach, integrating qualitative data from semi-structured interviews with school administrators, while quantitative data collected by questionnaire administered to teachers. Key themes that emerged highlight the significance of a growth-oriented educational atmosphere, proficient communication, and the incorporation of real-world applications in learning activities. Administrators emphasized the important role of mentoring, professional development, and explicit goal setting in fostering effective teaching methods. Nonetheless, they also acknowledged considerable challenges, including reluctance to change, constrained resources, and time limitations. Research findings indicate that innovative leadership tactics enhance student engagement, critical thinking, and cooperation, however their effect on standardized academic outcomes is more complex. The study advocates for a more systematic approach to leadership that include focused professional development, improved communication, and assistance for gradual transformation. Moreover, cultivating a conducive atmosphere that promotes risk-taking and cooperation is vital. The study enhances the current research by offering insights into effective leadership tactics amid changing educational requirements. It provides practical suggestions for educational leaders aiming to implement innovation sustainably in accordance with educational objectives. Keywords: innovative leadership, student achievement, educational strategies, professional development, teacher collaboration, technology integration, school management, growth culture, educational outcome
  • Thumbnail Image
    Item type:Item,
    Mentoring Programs and the Experiences of Teachers in UAE Primary and Secondary Schools
    (The British University in Dubai, 2025-06) SABIH, IMAN SOUD AHMAD; Dr Abubakar, Ahmed
    This mixed-methods study explores the structure, implementation, and impact of mentoring programs in primary and secondary schools in Abu Dhabi, United Arab Emirates. The research examines the experiences of mentors, mentees, and school leaders to understand how mentoring affects instructional effectiveness, professional growth, and teacher satisfaction. Drawing on theoretical frameworks including Social Learning Theory, Adult Learning Theory, and Cognitive Apprenticeship, the study investigates the types of mentoring programs used, the role of leadership in supporting mentoring, and perceived benefits and challenges of these initiatives. Quantitative data were gathered through a structured survey of 70 participants (40 mentors and 30 mentees), while qualitative insights were collected through semi-structured interviews with 14 stakeholders, including teachers and administrators. Findings indicate that mentoring programs are viewed positively overall, contributing to improved teaching practices, professional confidence, and collegial support. However, challenges such as time constraints, inconsistent mentor preparation, and mismatched expectations between mentors and mentees limit program effectiveness. The study offers practical recommendations for improving mentoring programs, including enhanced mentor training, institutional support, and structured feedback mechanisms. It concludes by emphasizing the need for culturally responsive, well-resourced, and flexible mentoring frameworks that align with the UAE's educational goals and diverse teaching environments. Keywords: mentoring programs; teacher development; professional growth; uae education; mentor-mentee relationship; instructional effectiveness; teacher retention; school leadership; mixed-methods research; educational reform