22505 research outputs found
Sort by
Mapping Accessibility in Dubai: Evaluating Al Wasl and Mirdif Against 20-Minute City Principles and the Dubai 2040 Urban Master Plan
No full textThe 15-minute city is a planning concept that aims to give people easy access to daily services such as schools, shops, healthcare, and recreation within a short walk or bike ride. Dubai has adapted this concept into a 20-minute city model under the Dubai 2040 Urban Master Plan. The goal is to reduce car use, improve quality of life, and create connected and sustainable neighborhoods. This study evaluates how well two neighborhoods in Dubai, Al Wasl and Mirdif, match the 20-minute city model. It analyzes three main goals from the Urban Master Plan: having 55% of residents within 800 meters of public transport, making 80% of essential services reachable within 20 minutes, and supporting walking and cycling through better transport networks. The study uses spatial analysis tools, travel time maps, and access scores to measure how easy it is to reach services and transit in each area. The results show that Al Wasl meets the 20-minute city goals across walking, cycling, and public transport because of its compact layout, metro access, and mix of land uses. Mirdif does not meet the same targets, mainly due to car dependence, spread-out housing, and no metro station nearby. Even with the planned Blue Line metro station, only 38% of residents will live within 800 meters of transit, which is still below the 55% target. The research explores how Dubai’s urban form, transportation systems, and land use patterns influence accessibility within the 20-minute city framework. It also suggests directions for future research, including applying this method to other districts across Dubai and tracking changes after planned infrastructure investments. This research shows that good public transport alone is not enough. To support the 20-minute city model, urban plans must also include closer service locations, better walking conditions, and more balanced land use. The findings give helpful ideas for planners and policy makers in Dubai and other growing cities
A High-Efficiency LDO Regulator with Adaptive PSRR and Transient Enhancements
No full textReliable power delivery is essential for all computing systems, particularly those operating with limited or constrained energy sources, such as batteries. Power management circuits must provide energy efficiency, stability, and resilience to disturbances to support accurate and consistent system performance. This thesis presents a high-efficiency low-dropout (LDO) regulator featuring adaptive power supply rejection ratio (PSRR) and transient response enhancement techniques. By leveraging adaptive analog design strategies, the proposed regulator dynamically boosts performance at high load currents, maintaining optimal efficiency across the full load range while circumventing key trade-offs inherent to conventional LDO architectures. The design incorporates several novel circuit techniques to improve overall performance relative to state-of-the-art solutions. Implemented in a 55 nm CMOS process, the regulator is validated through extensive simulation, layout, and silicon fabrication, with physical testing to follow upon chip delivery. Its utility is further demonstrated through integration in a time-domain neuromorphic system, highlighting its applicability in edge computing environments
What’s Coco Eating Today: Using Fragmented Time to Reduce Stress
Get PDFIn modern society, electronic mobile devices have become an integral part of life and entertainment. From a developmental perspective, the abundance of virtual content has brought significant changes in emotional, cognitive, and social aspects. On this background, the rise of the gaming industry has made \u27digital games\u27 a common form of entertainment. As games become a normalized part of daily life, the variety of game types has expanded greatly. This led me to consider: through interaction design and using games as a medium, is it possible to reduce users\u27 stress in a short time and provide positive emotional feedback? Based on this idea, I designed a collection game that can be played during fragmented moments of free time. In this game, the designer needs to balance the length of gameplay with how the game is played, looking for a simple way to interact without losing the fun. The main gameplay focuses on randomness and collecting elements, allowing players to enjoy the game experience. At the same time, visual elements like animations and character interactions help keep the game lively. These visuals attract and entertain users, helping them relax and reduce stress. I believe that the virtual reward system used in digital games can meet some of the users’ emotional needs, leading to positive emotional feedback
Towards Robust Deep Learning for Medical Imaging with Limited and Noisy Labeled Data
Get PDFDeep learning has emerged as a powerful tool in medical imaging, assisting healthcare professionals with several decision-making tasks, such as disease diagnosis, surgical intervention, and treatment planning. Supervised deep learning methods typically require large amounts of high-quality labeled data for training. However, acquiring high-quality labeled data in the medical domain is challenging due to factors such as the high cost of expert annotation and the presence of label noise, often due to inherent user/expert annotation variability. Models trained on limited or noisy labeled data suffer from poor performance due to overfitting, thus reducing their generalizability and trustworthiness for medical applications. During recent years, several works have been proposed to tackle the challenge of learning with limited and noisy labeled data in general machine learning. However, the complexity of medical data, including factors like subtle distinguishing features, imbalanced classes, and different imaging modalities, makes this challenge even more prevalent in the medical domain. This dissertation explores multiple approaches to overcome the issue of learning with limited and noisy labeled data for robust medical image applications. We first began by investigating the impact of class-dependent label noise on medical image classifiers to understand the effects when noisy and clean target classes are visually similar. We then introduced a framework to enhance robustness against noisy labels using self-supervised pretraining. As multiple factors influence learning with noisy labels in medical image classification, including the number of classes, dataset complexity, learning with noisy label methods, noise types, and self-supervised pretraining approaches, we conducted an in-depth study on the benefits of self-supervised pretraining in improving robustness against label noise across various datasets, taking these factors into account. Next, to tackle the challenge of limited labeled medical data, we proposed an active learning pipeline that leverages multimodal information to learn from limited labeled data, reducing annotation cost. Furthermore, we developed a robust framework for training with noisy labels in imbalanced medical image classification by separating noisy from clean labels and gradually relabeling some critical incorrect samples selected using active learning techniques. Our final contribution is aimed towards robust multimodal learning by addressing the issue of hallucination in Vision Language Models (VLMs). For this application, we created a vision-language medical dataset with hallucination-aware annotations and established initial benchmarks for VLMs, laying the groundwork for future research in medical applications
Shopping Quest: Transforming the shopping experience by creating a more immersive and enjoyable experience
Get PDFIn the post-pandemic era, this thesis looks into how gamification and augmented reality (AR) can get customers back into physical retail settings. Brick-and-mortar stores have found it difficult to stay relevant as consumer preferences have shifted toward online convenience. Shopping Quest attempts to make in-store grocery shopping an engaging experience. The project started with qualitative surveys using a human-centered design process to determine the main frustrations of shoppers, which were navigation and remembering what to buy. The final idea makes use of a wearable AR interface that directs users with a virtual rocket and promotes exploration with collectible digital petals. These characteristics provide progression, reduce decision fatigue, and make shopping feel purposeful. Iterative testing was used to create the final prototype, which was improved with user input and public showcase presentations. Visual design decisions were informed by principles of spatial clarity and emotional resonance, borrowing from familiar digital languages and narrative cues. The platform repositions physical shopping as an emotionally rich experience by combining storytelling, navigation, and sensory interaction. This thesis provides a model for how retail environments can move beyond efficiency toward experiential value and shows how AR and gamified design can reengage users with real-world spaces
Predicting and Forecasting University Rankings in the UAE using Machine Learning
Get PDFHigher Education Institutions (HEIs) play a vital role in advancing knowledge economies, and institutional rankings are increasingly used as global benchmarks of academic performance and reputation. In the context of the United Arab Emirates (UAE), enhancing institutional competitiveness in global rankings aligns with national strategies such as UAE Vision 2030 and the Centennial Plan 2071. This research applies machine learning (ML) techniques to develop predictive models that forecast institutional ranking outcomes, enabling data-driven planning and continuous academic improvement. The study utilizes a dataset compiled from the QS World University Rankings (2020–2024), cross- referenced against the institutional listings maintained by the Commission for Academic Accreditation (CAA) to ensure alignment with the UAE’s accredited higher education landscape. It includes over 4,000 records covering performance metrics such as research output, faculty- student ratios, academic and employer reputation, international collaborations, sustainability, and graduate employability. Preprocessing steps included normalization, handling of missing values, and feature selection to improve modeling accuracy and robustness. Five machine learning algorithms were implemented in this study: XGBoost Tree, Neural Network, Linear Support Vector Machine (LSVM), Linear Regression, and Generalized Linear Model (GLM), to analyze institutional performance and predict global ranking placement. Among them, the XGBoost model achieved the highest predictive performance, reaching an correlation of 95%, indicating strong reliability in capturing institutional ranking dynamics. The evaluation focused on correlation coefficients, MSLE (Mean Squared Logarithmic Error), and feature importance analysis, providing a comprehensive assessment of predictive accuracy and the relative influence of institutional variables. To support future institutional planning, this study proposes the use of a dynamic dashboard that could visualize ranking trends and simulate the impact of performance indicators on predicted outcomes. Such a tool would enable university leaders to make timely, data-informed decisions and strengthen alignment with international benchmarks. This research highlights the potential of integrating machine learning into higher education strategy. By leveraging historical ranking data and performance indicators, the proposed approach supports predictive decision-making and long-term academic advancement. It is particularly relevant to UAE universities seeking to reinforce their international standing and fulfill national education and innovation goals. Future research may expand the framework to incorporate real-time institutional data, apply it to multiple global ranking systems, and extend comparative models across regional higher education landscapes. The findings demonstrate how machine learning can serve as a powerful enabler of foresight, excellence, and strategic planning in the higher education sector
Observational Predictions for Convective Common Envelopes
No full textCommon envelopes (CEs) are thought to be the main method for producing tight binary systems in the universe, as the orbital period shrinks by several orders of magnitude during this phase. Despite their importance for many stellar evolution channels, direct detections are rare, and thus observational constraints on common envelope physics are often inferred from post-CE populations. Recently, galactic population observations suggest that the CE phase must be highly inefficient at using orbital energy to drive envelope ejection for low-mass systems and highly efficient for high-mass systems. Such a dichotomy has been explained by an interplay between convection, radiation, and orbital decay. If convective transport to the surface occurs faster than the orbit decays, the CE self-regulates and radiatively cools. Once the orbit shrinks such that convective transport is slow compared to orbital decay, a burst occurs as the release of orbital energy can be far in excess of that required to unbind the envelope. With the anticipation of first light for the Rubin Observatory, we calculate observable quantities for convective common envelopes. In particular, for low-mass systems, we produce observables, including light curves and apparent magnitudes for the Rubin filters. For high-mass systems, we identify binaries where convective effects are important and calculate corresponding light curves. We explore convective CE envelope models for two candidate systems, M101 OT2015-1 and OGLE-2002-BLG-360, where convection provides a reasonable explanation for the observational data. In general, convection imparts a distinct, long-term signature in the light curves, which should be detectable with upcoming transient surveys
Relativistic Gas Accretion onto Supermassive Black Hole Binaries from Inspiral Through Merger
No full textAccreting supermassive black hole binaries are powerful multimessenger sources emitting both gravitational and electromagnetic (EM) radiation. Understanding the accretion dynamics of these systems and predicting their distinctive EM signals is crucial to informing and guiding upcoming efforts aimed at detecting gravitational waves produced by these binaries. To this end, accurate numerical modeling is required to describe both the spacetime and the magnetized gas around the black holes. In this work, we present two key advances in this field of research. First, we have developed a novel 3D general relativistic magnetohydrodynamics (GRMHD) framework that combines multiple numerical codes to simulate the inspiral and merger of supermassive black hole binaries starting from realistic initial data and running all the way through merger. Throughout the evolution, we adopt a simple but functional prescription to account for gas cooling through photon emission. Next, we have applied our new computational method to follow the time evolution of circular, equal-mass black hole binaries with different black hole spin configurations for ~200 orbits, starting from a separation of 20 gravitational radii and reaching the post-merger evolutionary stage of the system. We illustrate the spin-induced differences in the structure of the minidisks orbiting each black hole during the early inspiral. We show how mass continues to flow toward the binary even after the binary decouples from its surrounding disk, but the accretion rate onto the black holes diminishes. We identify how the minidisks are slowly drained and eventually dissolve as the binary compresses. We confirm previous findings that the system\u27s luminosity decreases by a factor of a few during inspiral; however, we observe an abrupt increase by ~50--100% (depending on the binary\u27s spin setup) in this quantity at the time of merger, likely accompanied by an equally abrupt change in spectrum. We demonstrate that during the inspiral, fluid ram pressure regulates the fraction of the magnetic flux transported to the binary that attaches to the black holes\u27 horizons. Finally, we explore the spin-dependent dynamics of jet launching and jet-jet interaction and discuss the potentially associated electromagnetic signatures