DR-NTU (Digital Repository of NTU)
Not a member yet
116018 research outputs found
Sort by
Who does singlehood best? A latent profile analysis
No full textResearchers have begun to address heterogeneity in the single population. Historically, singles have been differentiated either by their marital history (e.g., widowed, divorced, never-married) or by whether they considered their singlehood voluntary. However, these approaches leave many unanswered questions about heterogeneity among singles. Addressing heterogeneity is important in light of recent interest in understanding how well-being varies across categories of singles. In the present study, we address this question by using Latent Profile Analysis (LPA) on a large cross-sectional sample of American singles ( = 4,835). Using LPA, we identified nine distinct profiles of singles differentiated by their romantic relationship goals (partner-seekers, casual-daters, and non-daters) and the perceived pressure they report experiencing from others to find a romantic partner (low-pressure, medium-pressure, and high-pressure). We then compared the profiles on levels of well-being, ill-being, personality, and socialization, as well as demographic, socioeconomic, and relationship history variables. Our analysis reveals several patterns of cross-category variability in outcomes, including some outcomes that vary primarily by romantic relationship goals, some that vary primarily by pressure, and others that vary according to a combination of both. Our results also suggest that two groups (low-pressure casual-daters and low-pressure non-daters) fare particularly well. They are also relatively common, comprising 30% of the sample. A third group-high-pressure non-daters-fare very poorly, and though they comprise only a small proportion of the sample (1.5%) the implications of this group are substantial when considered at national scale. The implications of these findings for research on singlehood and well-being are discussed.Published versionThe author(s) declare that no financial support was received for the research and/or publication of this article
Spectroscopic data processing of 2D microspheres for machine learning applications
No full textThis study focuses on the development and characterization of core-shell
nanostructures, specifically utilizing microfluidic techniques and Raman
spectroscopy. The aim of this work was to optimize the process of fabricating micro
and nanospheres, particularly those made from two-dimensional (2D) materials, and
to prepare the data for subsequent machine learning applications. Due to time
constraints, while machine learning integration was not fully implemented, data preprocessing techniques, including noise removal, baseline correction, and spectral
fitting, were performed. These pre-processing steps were executed using various
spectral fitting methods (Gaussian, Lorentzian, Voigt, and GL-mixed) to select the
most appropriate model for the data. Additionally, a region of interest (ROI) selection
process was implemented, allowing users to focus on specific spectral regions for
analysis. Key features, such as peak intensity ratios, full width at half maximum
(FWHM), and center wavenumber, were extracted and exported for further analysis.
The use of microfluidics in combination with Raman spectroscopy offered insights
into the structural and compositional characteristics of the nanomaterials. This work
serves as a foundation for future machine learning integration to automate the
characterization and optimization of core-shell nanostructures based on spectral data,
which can ultimately advance the understanding of their properties and applications in
biomedicine, energy storage, and electronics.Bachelor's degre
GNSS signal quality monitoring
No full textGlobal Navigation Satellite System(GNSS) signal quality monitoring has been
implemented worldwide. With the increase in GNSS spoofing attacks each year, it is
critical to be able to mitigate or even prevent these situations from happening as the
results could be impactful to the individual.
In this project, data collection methods utilising the HackRF device as well as the
smartphone Samsung Galaxy S21 and S22 device was being conducted. The data
logging process on the Samsung Galaxy S21 and S22 device was conducted using
the application “GNSSlogger” found in the android store. The data collection process
did not manage to achieve the expected results when taking reference to the NASA
dataset and results were not optimal. The ranges found from the data collection
process was between 390m to 510m while the expected range was between
20,000km to 30,000km.
Datasets from Texas Spoofing Test Battery (TEXBAT) were being used later in the
project to further test GNSS signal quality, specifically the clean dataset and the DS3
spoofed dataset. DS3 dataset contains data collected from a static power push
scenario. Through the comparison of the datasets, spoofing attempts were being
identified through machine learning models such as random forest, isolation forest
and one-class SVM to detect any anomalies and differences in the dataset. The
results derived from this experiment have shown positive results and presence of
spoofing being injected into the signal.
Despite the multiple challenges and attempts in data collection at the start, the use of
TEXBAT data has shown positive results, showing signs of spoofing through
navigation coordinates as well as power difference in the dataset. With the research
and experiments conducted, further implementations could be achieved through other
detection methods after collection of raw data in an open environment with minimal
interference.Bachelor's degre
Context-aware techniques for real-time decision making in AMR (path planning)
No full textClassical planners like A* are widely used in global path planning. However, classical
planners struggle in dynamic environments, where the global path is unable to account for
unexpected obstacles. This poses a challenge for Autonomous Mobile Robots (AMRs) that
rely on efficient path planning and obstacle avoidance. Deep Reinforcement Learning (deep
RL) has the ability to learn from its environment and make decisions autonomously, which
highlights it as a potential solution for dynamic path planning. Building on a deep RL
foundation, we developed Smart Algorithm Switching (SAS), a hybrid approach that
integrates Soft Actor-Critic and A* to perform real-time path planning and obstacle
avoidance. Our experimental results demonstrate that SAS performs better at obstacle
avoidance and goal-seeking behavior than SAC alone.Bachelor's degre
Romantic love TM — terms and conditions apply
No full textWhat makes romantic love unique as a category of its own—distinct from friendship
and other non-romantic intimacies? Existing definitions of romantic love are often either too
narrow (thereby excluding variants of romantic love’s practices today) or too broad (applying
also to friendship and non-romantic intimacies). These limitations highlight the need for a
dialectical approach that considers tensions between romantic love’s conventional
constitution and its potential for evolution. This dissertation reconceptualizes romantic love’s
core through the lens of social normativity to account for this dialectical tension. Examining
experiences of dissonance and discomfort, I innovatively examine how romantic love is
presently maintained as coherent social category through its ideological mechanisms.
To do this, I rearranged ideas from existing definitions and refined these through
interviews with 45 young adults in Singapore spanning various sexuality and gender
identities, including the aromantic and asexual spectra. Using an interdisciplinary method of
distillation to find what seem to be most definitive of romantic normativity specific to today’s
global world-historical conditions, this research identifies key romantic templates. These
templates are ideal types—conceptual constructs that accentuate how romantic love works
through aspirational ideals. These identified templates provide a scaffold for near-future
research on romantic love, especially timely given the increasing fetishization of romantic
love as an autonomous sphere of symbolic normativity under widespread neoliberal
commodity-thinking and globalized information society.
If romantic love were a bundle of social norms, what would they be? I posit that
romantic love today comprises three overarching normative templates (Passion,
Commitment, and Compulsory Sexuality) that govern feelings, relationship expectations, and
sexual practices respectively, with each containing sub-templates. Working closely in tandem,
these templates do not only instruct people on how to live out “authentic” romantic intimacies
but also serve a higher-order function: maintaining romantic love’s distinction from and
superiority over other genres of intimacy. Nonetheless, people can and do accept, modify, and
reject romantic sub-/templates in myriad combinations, which need not align with how they
label their identities and experiences. As a complementary concept, I define “doing romantic
love” as social practices that reinscribe romantic normativity.
Consequently, this dissertation redefines what it means to be aromantic. I coin the
term alter-romantic to highlight diverse ways through which people manifest alter-romantic
tendencies—through their rejection or modification of templates, or labels, or both. I define
alter-romanticism as an experiential vantage point of dissonance from which the assumed
coherence and naturalness of romantic love templates and/or labels may be questioned,
necessarily involving the ongoing struggle and negotiation with romantically normative
demands. Aromantic-identifying people are therefore a subset of an umbrella of people who
experience discomfort towards romantic love. Alter-romanticism accounts for those who
practise romantic love divergently, but do not identify as aromantic. This highlights the
limitations of scalar and one-dimensional understandings of aromanticism and addresses the
major lack of research on aromantic experiences as a phenomenon independent of asexuality.
This dissertation’s insights are likely to resonate across globalized urban societies
today and have transformative implications for psychotherapy, coaching, and education.Doctor of Philosoph
Design of a fully differential controlled gain open-loop wideband amplifier with high PSRR
No full textThis project shows the design process of a fully differential controlled gain open-loop wideband amplifier with high PSRR on 0.55nm CMOS technology. The amplifier uses a 1.8V voltage supply with a current consumption of 723uA at 1.301mW. The amplifier achieves an open loop gain of 12.56dB at 1kHz. The Unity Gain bandwidth of the amplifier is 323MHz with 100fF load on both outputs. Output voltage can swing from 0.215V to 1.675V. An RTO PSRR of 41dB and CMRR of 65dB. The circuit has a Total Harmonic Distortion of -51.2dB. Monte Carlo Simulation is done to verify the effects of transistor mismatch.Bachelor's degre
A review on liner shipping companies' investment strategy in the logistics sector (Hapag-Lloyd)
No full textThis study evaluates Hapag-Lloyd’s "Pure Play Plus" investment strategy in the logistics sector,
contrasting it with the full-scale vertical integration models of Maersk and CMA CGM. Using a mixedmethods approach—comprising surveys, expert interviews, and secondary data analysis—the research
investigates the financial, operational, and strategic drivers behind Hapag-Lloyd’s selective acquisitions
of logistically enabled terminals.
The findings reveal that Hapag-Lloyd’s strategy delivers measurable short-term financial benefits,
notably EBITDA margin improvements within two quarters post-acquisition of mature, revenuegenerating assets such as SM SAAM S.A. and J M Baxi. Operationally, the integration of inland
transport, warehousing, and value-added services enhances connectivity and service reliability and
reduces reliance on third parties, aligning with RBV and Core Competence Theory. Strategically, the
tapered integration model mitigates the higher competitive pressures faced by full logistics integrators,
as demonstrated through Porter’s Five Forces analysis.
However, the study identifies risks associated with the Partial Equity Model (PEM), including gradual
leverage accumulation and liquidity erosion due to successive acquisitions. Despite these risks, HapagLloyd’s disciplined, infrastructure-centric approach offers a sustainable balance between growth,
financial prudence, and operational control, making it suitable for risk-averse carriers.
Key recommendations include enhancing financial transparency through activity-based costing,
standardising operational synergy tracking via key performance indicators (KPIs), and targeting
terminals with embedded, asset-light, digitally enabled logistics solutions to improve flexibility and
customer responsiveness.
In conclusion, Hapag-Lloyd’s selective integration strategy effectively balances short-term profitability
and operational efficiency while avoiding the complexities of full-scale logistics integration. To sustain
long-term competitiveness amidst evolving digital and customer-centric logistics demands, the
company must complement its asset-heavy investments with scalable, technology-driven solutionsBachelor's degre
Implementation of a discrete-time signal calculator using MATLAB app designer
No full textThis final year report focuses on designing and developing a graphical user interface to
implement a Discrete-Time Signal Calculator using MATLAB’s App Designer, providing ease
of learning for students learning EE2010 Signals and Systems. This project aims to provide
students with a friendlier and more intuitive way to understand and learn discrete-time signals,
which are fundamental in the study of digital signals. This will help provide better visualization
for students to see the concepts used in EE2010 such as correlation and convolution, which will
help deepen their understanding of different concepts.
The implementation of this project will be done with MATLAB Designer, which will be used to
create the calculator with its application of the GUI aspects. MATLAB is also being used to run
the codes. The purpose of using MATLAB is due to the many different functions within the
software for development.Bachelor's degre
Assessing a decision-support tool to estimate the cooling potential and economic savings from urban vegetation in Singapore
No full textTo mitigate the growing threat of urban heat, cities are implementing greening strategies such as tree planting or the development of parks. Effectively integrating these solutions into planning requires quantitative information on the cooling effect of urban vegetation. Here we examined the performance of an open-source decision-support tool, the Integrated Valuation of Ecosystem Services and Tradeoffs Urban Cooling model, to estimate the cooling effect and economic benefits from urban vegetation in a tropical city context, using Singapore as an exemplar case study. Using observed temperature data, we calibrated the model to estimate the spatial distribution of annual average day- and night-time temperature at 10 m spatial resolution and validated the results using leave-one-out cross validation. The calibrated models performed well to estimate annual average daily mean and maximum (day), and minimum (night) temperatures (R2 of 0.78, 0.65, and 0.52, respectively). We estimated that urban cooling in Singapore provides economic savings of $47.14 million SGD annually from reduced energy consumption in public residential buildings, based on the relationship between energy consumption and mean temperature. Our results give confidence in the model as a decision-support tool to estimate urban heat island effects and evaluate heat mitigation strategies in tropical cities.National Research Foundation (NRF)Published versionThis research was funded by the National Research Foundation of Singapore, Prime Minister’s Office (NRF-NRFF12-2020-0009; Cooling Singapore 2.0 Funding Initiative)
Model a 6-DOF robot manipulator and its working environment, and design and simulate a tracking control method for the robot
No full textThis paper focuses on the modelling and control simulation of a 6 Degrees of Freedom (DOF)
robotic manipulator, focusing on its ability to follow a predefined trajectory with high precision
and reliability. Robots are widely recognized for their efficiency, resilience, and precision,
making them indispensable in industries where consistency and repeatability are critical. The
main objective of this project is to model a 6 DOF robot manipulator and its working
environment, and to design and simulate a tracking control method that enables smooth and
accurate motion. The mechanical structure of the robot was developed using SolidWorks and
imported into Simulink via the Simscape Multibody environment to allow dynamic simulation.
A PID controller is applied at each joint to regulate motion, while both forward and inverse
kinematics are implemented using MATLAB scripts to compute joint angles and end-effector
positions. The result is a working simulation that demonstrates the robot’s ability to follow
desired paths realistically, showcasing the potential of simulation-driven design for robotic
systems.Bachelor's degre