United Arab Emirates University
United Arab Emirates University: Scholarworks@UAEU / جامعة الاماراتNot a member yet
5830 research outputs found
Sort by
COLLABORATIVE NETWORK TRAFFIC MANAGEMENT STRATEGIES USING THE DISTRIBUTED REINFORCEMENT LEARNING AND LARGE LANGUAGE MODELS
Full text linkThe focus of this research is to explore collaborative network traffic management strategies using the Distributed Reinforcement Learning (DRL) and Large Language Models (LLMs) approaches. It emphasizes exploring a new tool for addressing network traffic by utilizing Distributed Reinforcement Learning (DRL) and Large Language Models (LLMs). This is achieved by utilizing self-organizing and self-directing techniques to optimize the network performance. Using the NF-TON-IOT dataset, various classifiers such as Random Forest, AdaBoost, C4. 5, Multi-Layer Perceptron (MLP), and SVM with an RBF kernel were tested for traffic classification and intrusion detection. Research recommends that DRL optimizes the complexity of the network by allowing agents to make decisions independently of the other agents; LLM optimizes the interaction between the agents in the network. By using the performance analysis, it has been explored that Random Forest and AdaBoost are more effective as compared to other classifiers that have been tested such as SVM with RBF kernel. However, the RBF kernel SVM has the drawbacks. The main drawback of the SVM RBF kernel is associated with its computational expenses and the longer training time especially in datasets of sizeable amount as NF-TON-IOT. This is not scalable and inefficient for real-time network traffic management scenario since fast adaptive responses are further needed.The overall implementation of different confusion matrices for the accuracy check demonstrates that combining the DRL and LLM frameworks is conceivable for constructing a flexible and extensible network management architecture that can provide valuable guidance for the future development of network traffic management
DESIGN AND EVALUATION OF GEOPOLYMER COMPOSITE AS A POTENTIAL SORBENT OF HEAVY METALS FROM WATER
Full text linkReducing the levels of heavy metals in water is crucial, given their severe environmental and health impacts. Various methods exist for heavy metals removal. Yet, they mostly come with multiple drawbacks, such as costly treatments and limited efficiency. Previous studies have found geopolymer to be a promising sorbent because it is synthesized using by-product materials, making it an eco-friendly and economically sustainable alternative. Limited studies explored the sorption potential of fly ash-slag blended geopolymers; none examined the effect of mix design factors synergically on sorption, mechanical properties, and durability, and few considered solution characteristics and operating conditions simultaneously. Therefore, this thesis aims to develop and evaluate a fly ash-slag blended geopolymer sorbent for heavy metals removal from wastewater under varying conditions. The work was implemented through two main phases. In the first phase, an initial study was conducted to design 16 geopolymer mixes and select the optimum mix. In phase two, the impact of various parameters on the lead (Pb2+) uptake capacity of the selected optimum geopolymer mix was examined. In phase one, single and binary fly ash and slag blends were utilized as the binding material. The geopolymer composite was activated using either sodium hydroxide solely or in combination with sodium silicate. The Taguchi method was employed to design geopolymer mixes, having four factors, each with four levels of variation. These factors included the fly ash-to-slag ratio (FA), binder content (BC), the molarity of the sodium hydroxide solution (M), and the sodium silicate-to-sodium hydroxide ratio (SS/SH). The performance of the geopolymer sorbent was rigorously assessed against a comprehensive set of responses categorized into synthesis and performance criteria. The TOPSIS methodology was applied to aggregate the response criteria and determine the optimal mix for superior performance. A sensitivity analysis was performed to study the sensitivity of the results to the weights assigned for each criterion. In this phase, the results showed that the optimum mix consisted of an FA of 33%, BC of 1050 kg/m3, M of 10, and SS/SH of 3. Phase two investigated the impact of various parameters on the Pb2+ uptake capacity of the selected optimal geopolymer mix. These parameters include changes in solution characteristics and variations in operational conditions. Furthermore, the impact of introducing a foaming agent to the optimum mix was examined. The results demonstrated that the removal efficiency increases with increasing geopolymer dosage, contact time, temperature, and the decrease of geopolymer particle size and Pb2+ initial concentration. Moreover, it has been observed that adding a foaming agent to the geopolymer mix enhances the removal efficiency. The optimum removal efficiency was obtained at a final pH of 5. The kinetic data were found to fit the pseudo-second-order kinetic model. Also, the sorption isotherm study indicated that the experimental data showed a high nonlinearity, and the Langmuir model fits the data better than the Freundlich model. This study demonstrated the potential of using geopolymer as a sorbent in removing heavy metals from water, addressing a critical environmental concern with great implications for practical applications. Future studies should focus on investigating the performance of geopolymer composites in large-scale productions and industrial real-life wastewater instead of synthetic wastewater. Additionally, further exploration into the valorization and regeneration of geopolymer composites and sustainable final disposal strategies should be performed. Moreover, expanding the environmental and cost-benefit analyses conducted in this study by including a life cycle assessment to evaluate the geopolymer performance as a sorbent compared to traditional sorbents is recommended
DEVELOPMENT OF AL AIN FLASH FLOOD RISK MAPS UTILIZING HIGH-RESOLUTION AND ACCURATE REMOTE SENSING DATA
Full text linkAl Ain City has undergone significant land use transformations from 1992 to 2022, marked by a population increase of over 50% and extensive development activities. This period has seen the construction of new residential areas, infrastructure, and commercial establishments, as well as enhancements to green spaces. An analysis of the Digital Terrain Model (DTM) revealed an 83% alteration in the terrain due to urbanization and shifting sand dunes, with most elevation changes under 5 meters. The primary aim of this dissertation is to develop accurate flash flood risk maps for Al Ain using high-resolution remote sensing data. This research highlights the dramatic changes in the city’s terrain over the past three decades and emphasizes the need for maintaining an up-to-date DTM. Airborne LiDAR technology is proposed as the most effective method for generating precise DTMs, while high-resolution photogrammetry (e.g., 10cm resolution) is identified as a secondary option. The dissertation critiques the use of low-resolution open-source Digital Elevation Models (DEMs) like DEM-SRTM, which are inadequate for detailed urban flood risk assessments. To achieve these objectives, the methodology involved a literature review, acquisition of high-resolution remote sensing data, and the application of advanced processing techniques to create accurate DTMs. Historical terrain data from 1992 and 2022 were analyzed using change detection methods and GIS analysis. Flash flood risk maps were developed based on processed DTMs, identifying varying risk levels across the landscape. Preliminary findings indicate substantial land changes in Al Ain, with 257 km² transformed into built environments and green spaces. The study reveals that the total stream length increased from 221.451 km in 1992 to 2904.1 km by 2022, demonstrating significant environmental shifts. The project also emphasized the importance of accurate DTMs in flood risk management, prompting Al Ain Municipality to commit to regular updates of high-resolution geospatial data. In 2022, a project was initiated to enhance the DTM for Al Ain, involving high-resolution imagery and dense LiDAR point clouds. This collaborative effort aims to create accurate digital terrain and surface models, ultimately aiding in the development of reliable flash flood risk maps. The Municipality has agreed to allocate a budget for biennial updates, ensuring that geospatial resources support ongoing analysis and risk assessments. These findings underscore the necessity for continuous monitoring and assessment of Al Ain’s evolving landscape to enhance flood risk management strategies
Utilizing Infographics to Enhance Some Continuous Learning Skills for the Educator to Keep Pace with Labor Market Developments Using Electronic Resources
Full text linkThe research aimed to identify the Effectiveness of employing infographics to develop some continuous learning skills among female student teachers. The sample consisted of (60) second-year students at the Faculty of Early Childhood Education, divided into two groups: an experimental group (30) and a control group (30). The research followed the experimental method, and the researcher used a scale to measure continuous learning skills using electronic resources for the student teachers, along with a training program that employed infographics to develop continuous learning skills. The application lasted for (two months).
The results revealed that there were statistically significant differences between the mean scores of the female student teachers in favor of the post-test results of the experimental group.
Keywords: Infographics - Continuous learning skills - Electronic resources
Exploring the Relationship between Students\u27 Attitudes toward Mathematics and Mathematics Achievement in the UAE
Full text linkThis study presents the relationship between students\u27 attitudes toward mathematics and mathematics achievement in the United Arab Emirates (UAE) using the Trends in International Mathematics and Science Study (TIMSS) 2019 data. The target population is grade 8 UAE students, and the sample consisted of 22,334 eighth-grade students. The researcher used the variables like learning mathematics, confident in mathematics, and value mathematics to represent the independent variable, students\u27 attitudes toward mathematics. The dependent variable, achievement in mathematics, contains four content domains (algebra, geometry, data probability, and numbers) and three cognitive domains (knowing, applying, and reasoning). After conducting regression analysis, the researcher found a positive relationship between students\u27 attitudes toward mathematics and mathematics achievement. Both confidence in mathematics and valuing mathematics significantly predict mathematics achievement, while there is no significant relationship between liking to learn mathematics and mathematics achievement. The study recommends using instructional strategies to improve students’ attitudes toward mathematics, such as constructive feedback and solving real-world problems.
Keywords: attitudes toward mathematics, mathematics achievement, grade 8, TIMSS
CATALYTIC HYDRODEOXYGENATION OF BIOMASS MODEL COMPOUNDS USING NI, CO AND FE OVER HZSM-5
Full text linkWith the exponential increase in the human population and energy demand coupled with the alarming levels of worldwide CO2 emissions and the global warming crisis, there is a shift of focus towards developing and deploying clean and renewable energy sources. Biomass is a widely available renewable energy source that offers a promising sustainable alternative to fossil fuels, especially in the case of liquid biofuels production. However, unlike fossil fuels, biomass inherently contains high oxygen content which must be reduced in the biomass-to-biofuel upgrading process in order for the biofuel to provide comparable amounts of energy per unit volume to traditional fuels currently used as well as matching their physicochemical properties. Therefore, hydrodeoxygenation (HDO) is an essential step lying at the heart of the upgrading process. This research project aims at upgrading biomass model compounds, namely vanillin and guaiacol, into partially deoxygenated biofuel precursor through the hydrodeoxygenation reaction. The deoxygenation pathway investigated starts with carrying out a hydrodeoxygenation reaction on vanillin to produce guaiacol, followed by another hydrodeoxygenation reaction on guaiacol to produce phenol. Novel catalysts are prepared and tested at different metal loadings, and the performance of the catalysts is evaluated based on the conversion, selectivity, yield and the degree of deactivation. In this work, the synthesized catalysts use nickel (Ni), cobalt (Co) and iron (Fe) and the noble metal rhodium (Rh) supported each on the HZSM-5 zeolite. Various characterization techniques were used, including Fourier Transform Infrared Spectroscopy (FTIR), Thermogravimetric Analysis (TGA), X-ray Diffraction (XRD), Temperature Programmed Reduction (TPR) and Scanning Electron Microscope (SEM), to elucidate the catalysts’ structures and to comprehend the origins of the observed catalytic activity in the hydrodeoxygenation. Seven hydrodeoxygenation reactions were carried out on vanillin with H2 flow rate of 100mL/min at 300℃ and atmospheric pressure. Ni/HZSM-5 catalyst showed the best performance among the transition metals with the peak performance being at the 10% loading with a conversion of 65.64% and a 96.35% selectivity towards guaiacol. Loading the same catalyst with 0.5% of the noble metal rhodium (i.e. 0.5%Rh-10%Ni/HZSM-5) boosted the conversion to 72.86% while maintaining almost the same selectivity towards guaiacol. Several products were obtained from the guaiacol hydrodeoxygenation at the same conditions over 10%Ni/HZSM-5 including phenol, m-cresol and o-cresol, with phenol dominating at 57.5% selectivity. The ultimate goal of the project is to optimize the hydrodeoxygenation reaction for biomass upgrading through developing novel and efficient catalysts as well as exploring a new practical reaction pathway. Achieving this goal will make biofuels more obtainable and generate greater interest in this type of renewable energy sources, which will eventually reduce the environmental damages caused by the conventional fuels that have been powering transportation for decades
NEW CONTRAST-BASED METRICS FOR ROADWAY LIGHTING DESIGN
Full text linkRoadway lighting standards specify design targets of horizontal illuminance or luminance at the roadway surface for various roadway types. Although the luminance and illuminance method to roadway lighting standards incorporate some factors that affect human visibility, other important factors, such as the target’s contrast and size, are not examined in these standards. Roadway lighting standards should consider how the human visual system detects objects and provides comfortable approaches to improve visibility for drivers. As luminous contrast has a crucial impact on visibility measures, recent studies have been directed to propose new approaches that utilize it. This research aims to propose new roadway design indexes based on the contrast of obstacles and to investigate whether they can be integrated with or replace the design based on luminance or illuminance. An experimental program was performed on a two-lane road located in Falaj Hazzah. Grids comprising five models, each model representing obstacles of different heights (0.05, 0.5, 1.0, 1.5 m), were distributed along the lateral and longitudinal directions of the road and systematically moved between two streetlights. A vehicle-mounted camera was placed at a distance equal to the total stopping distance from each obstacle to capture calibrated high-dynamic-range (HDR) luminance maps for diverse scenarios. HDR images provided detailed luminance data to evaluate factors influencing contrast. The dynamic of contrast results was investigated under various roadway lighting design variables, such as pole spacings, lumen output, and vehicle headlights. Dialux software was utilized to replicate the existing road to validate the model as a design tool. The contrast results were then used to propose two different metrics for roadway lighting standards, named the Useful Contrast Index (UCI) and the Mean Contrast Index (MCI). The performance of the proposed metrics was evaluated by changing lighting design variables such as lumen output and pole spacings. Results showed a weak correlation between obstacle contrast and road pavement luminance, highlighting the need to incorporate new contrast-based approaches into current roadway lighting design. For very low obstacles, contrast decreased with increased lumen output, whereas for taller obstacles (1.5 meters), contrast improved with higher lumen output. While reduced pole spacing increased pavement luminance, it did not increase the contrast of obstacles. The results also show that the proposed contrast-based indexes have merit and could be integrated into current roadway lighting guidance, aiming at enhancing the visual environment for drivers and improving safety
KNEE-THIGH-HIP IMPACT IN CHILDREN: EFFECTS OF AGE AND IMPACT VELOCITY
Full text linkBrief introduction: This thesis examined the crash impact injury outcomes of the knee-thigh-hip complex (KTH) in children based on age and impact velocity using computational modelling. Child occupant (3YO, 6YO, 10YO) Total Human Body Model for Safety (THUMS) models developed by Toyota were used to evaluate the effects of age and impact velocity on pediatric KTH crash impact response. Aims: The main objectives of this thesis are to analyze the effect of child age and impact velocity on KTH crash impact response. Methods: The knee impact simulations were performed based on different impact velocities (1.2, 3.5, 7.2, and 11.1 m/s). The knee was impacted at the middle of the knee joint, with the impactor hitting both the femur and tibia at the patella for all the three models. Simulation results were analyzed to evaluate the loading characteristics at the femur, tibia, knee, and ligaments across the 3 models. These characteristics include the peak force, absorbed energy, and peak stress. KTH loading characteristics were then evaluated for trends between the 3 different models at varying impact velocities. Results: The results of the simulations demonstrate that the child knee joint’s response to impact forces varied significantly based on the impact velocity and the age of the child model. The force-displacement diagrams provided valuable insights into how the child knee joint reacts under different impact velocities, showcasing a clear trend of increased force with higher impact velocities. These diagrams also highlighted that the child knee joint’s ability to absorb impact energy decreased with higher impact speeds which is critical for understanding injury mechanisms in real-world crash impact scenarios. The variations in force-displacement characteristics, energy absorption, and stress distributions across the three child models underscores the importance of accounting for age when evaluating knee crash impact injury risks in children. Significant contributions: This is the first study to evaluate the effects of age and impact velocity on pediatric KTH response. The findings of this study revealed that age-based intrinsic factors affect the impact response of the KTH complex in children. An understanding of child KTH crash impact injury mechanisms is important to reduce the burden of these injuries in children. Gap filled: The thesis demonstrates the importance of numerical simulations in evaluating pediatric knee joint biomechanical tolerance under crash impact conditions. The combination of various impact velocities, model age, and the analyses of key biomechanical variables provide a deeper understanding of how the child knee joint behaves under crash impact conditions thereby enhancing the assessment of KTH injuries sustained by children in vehicle crashes and their associated risk factors. These findings can be instrumental in improving vehicle safety standards and in designing vehicle crash protective systems for children
CONSTRUCTION OF STOCK PORTFOLIOS BY MACHINE LEARNING METHODS
Full text linkWe study the theory and application of two machine learning (ML) algorithms: Long-Short Term Memory Network (LSTM) and Random Forest. The study begins by providing an overview of mathematical foundation, highlighting the significance of mathematics in machine learning. We study and explain the mathematical details involved in these two algorithms. We also study closely related subjects which include but not limited to gradient descent, automatic differentiation, and recurrent neural network. The main ideas behind these topics form the foundation of any ML algorithms.
As an application of the ML algorithms, we focus on the U.S. stock market. We build stock portfolios by various strategies which are primarily based on LSTM and Random Forest. This thesis can be viewed as an application of machine learning to the field of stock market analysis. The stock portfolios are constructed based on certain degree of diversification across diverse industries in the U.S. economy and fixing a specified number of stocks throughout the backtesting period we considered. We use either technical indicators or fundamental indicators as the input features to the ML algorithms for each stock. We predict the stock price of each stock based on various strategies. We rank the stocks based on the predicted return in a given trading year and select the top stocks as our porfolios for the year. We do re-balancing in each year based on the same strategies for all the stock portfolios during the backtesting period.
We compare how these stock portfolios perform against market indices like S&P 500. The performance of a stock portfolio is measured by its compounded annual growth rate (CAGR). We check and analyze whether the chosen machine learning algorithms can be valuable tools for picking stocks based on historical stock data. We determine if the stock portfolios constructed could consistently outperform the market benchmark over the chosen backtesting period. We present these results and analyze the strategies in details. We discuss possibilities of extending the main ideas contained in this thesis for future work
A BAYESIAN BELIEF NETWORKS APPROACH TO PRIORITIZE ROAD SAFETY MITIGATION MEASURES FOR ECO-MOBILITY MODES
Full text linkThis study presents an innovative decision-making approach that enables the prioritization of road safety measures for eco-mobility users—pedestrians, cyclists and light-electric micromobility riders—in Abu Dhabi. The analysis uses 2032 eco-mobility crash records (1016 pedestrian, 567 bicycle, 449 e-scooter) from 2020-2023 supplied by Abu Dhabi Police and the Department of Municipalities and Transport. A supervised Bayesian Belief Network (BBN) was constructed around six empirically estimated variables: visibility, predictability, road curvature, obstructions, lighting and speed limit. Maximum-likelihood conditional-probability tables were derived from the cleaned dataset; Beta priors were applied only where cell counts were sparse. Ten conventional machine-learning models—including Random Forest, Gradient Boosting, SVM and LSTM—were trained on the same features for benchmarking. The BBN achieved the strongest overall results (precision = 0.999; recall = 0.998; R² = 0.998), outperforming all other models by at least four percentage points on F1-score. Scenario analysis showed that maintaining high visibility and high predictability lowers crash probability by 90% whereas sharp curvature and major obstructions more than triple risk. To translate these probabilistic findings into action, a cost-benefit module estimated benefit-cost ratios for 12 candidate counter-measures. The highest-ranking interventions were 30 km h⁻¹ traffic-calming packages in school and residential zones (median 15% crash reduction; BCR = 3.7) and targeted enforcement of pedestrian right-of-way laws (12.5% reduction; BCR = 2.9). The combined BBN–economic framework offers Abu Dhabi planners a transparent, data-driven tool for sequencing low-carbon mobility safety investments under budget constraints. Beyond the local case, the methodology can be adapted to other cities that aim to improve soft-mobility safety while progressing toward Vision-Zero objectives