NUI Maynooth Eprint Archive
Not a member yet
18159 research outputs found
Sort by
Artificial Intelligence The Human Autonomy and Ethical Considerations of Advancing Intelligent Systems and Machines
No full textIn the modern era, the Artificial Intelligence (AI) concept existed since Alan Turing conceived Machine Learning in 1935.1 AI, as a term, came into more common and public discourse in early 2023, when Elon Musk, industry leaders and others called for more regulation. This marked a turning point for more public discussion. Popular views range from, AI bringing enormous benefits, to AI representing an existential threat for mankind. Understanding the dichotomy between AI’s promise and its perils is vital for humankind, human dignity, autonomy and freedom. As a technology AI disrupts society and its norms. While acknowledging the benefits; there are moral and ethical issues, along with serious threats. To bring an understanding of AI itself, human intelligence and consciousness; as well as computer-based intelligence and consciousness must be considered. The moral standings of humans and AI must be evaluated and assessed if they are or could be equal. Should AI become superior to human intelligence, “Strong AI,” humans may become subservient to it, then fundamental human rights questions arise. Humans and AI companies maintain that they will always be in full control of Weak and Strong AI, however the reality might suggest the contrary, if the underlying ways of human life are altered. In its current iteration, Weak AI has and will continue to irreversibly change people’s behaviours, humans become reliant on AI to the degree of being unable to conduct ordinary aspects of everyday life in its absence. While multiple problems exist in the world, AI further harms human autonomy and is socially disruptive. ‘Big Data’, the ‘Internet of Things’ and ‘always connected’ systems, are part of AI and are driven by AI. Its power lies not just with these technologies, but also with the algorithms’ subliminal effects. The relatively limited number of Big-tech companies sway governments, using their enormous economic and political clout. This thesis sets out to demonstrate how, while Weak AI may not be a true existential threat to mankind; there are nevertheless serious moral and
ethical risks and dangers that overshadow its unquestionable benefits. However, strong governments can resolve these dilemmas
Exploring the Role of Pellino Proteins in Cells of the Adaptive Immune System
No full textThe innate and adaptive immune systems intertwine in mounting an effective immune response against invading pathogens. The activation of innate immune components is often a prerequisite for the initiation of the adaptive immune response. The Pellino family consists of a 3-membered family of E3 ubiquitin ligases (Pellino 1, Pellino 2, and Pellino 3) that play important roles in immunity by catalysing post-translational modification of important signalling molecules. Pellino proteins have been widely studied for their roles in innate immunity. Nevertheless, emerging reports have highlighted the potential of Pellino proteins in regulating adaptive immune system. This thesis aims to further the knowledge in this area by performing the first systematic characterisation of the role of Pellino proteins in generating adaptive immune cell populations. To this end novel genetic models were generated resulting in mice that lack individual and combination of the Pellino family. Such models allowed for the first time to investigate potential functional interactions between the Pellino family members. While Pellino 3 does not mediate the production of various T and B cell subsets, Pellino 2 has a role in CD4 and CD8 T cell activation that is dependent on age. Importantly, the findings also highlight a selective role for Pellino 1 in negatively regulating the generation of activated CD4 and CD8 T cells, as well as germinal centre B cells and plasma cells in both young (10-12 weeks old) and aged (6 months old) mice. Interestingly, Pellino 1 and Pellino 2 exhibit distinctive functional roles in mediating CD8 T cell activation as individual deficiency of Pellino 1 and Pellino 2 favours its differentiation into CD8+ effector memory (Tem) and central memory (Tcm) cells, respectively. This suggests that they might have different physiological roles in controlling cytotoxic functions of Tem cells or systemic infections through Tcm cells. It was also found that Pellino 1 negatively modulates IL-17 production in Th17 cells. The mechanistic basis to the role of Pellino 1 in controlling T cell activation and IL-17 production is also explored. The studies conclude that this regulatory function of Pellino 1 is intrinsic to T cells and not antigen presenting cells. Overall, this body of work provides novel insight into the role of Pellino proteins, particularly Pellino 1, in the adaptive immune system. It also forms a foundation for future research to elucidate the physiological role of Pellino 1 in Th17 cell differentiation and may represent a new pathway that may be open to therapeutic exploitation in the treatment of inflammatory diseases
Quantum algorithm for linear systems of equations for the multi-dimensional Black-Scholes equation
No full textThe primary focus of this thesis is the investigation of the quantum algorithm
for linear systems of equations (HHL) for the valuation of multi-asset options,
a particular type of financial instrument. Quantum computing has the possibility
to revolutionize many fields that are computationally intensive, such as
quantitative finance. We extend the previous works on quantum solutions to the
Black-Scholes equation for option pricing and provide its proof-of-principle implementation.
We transform the problem of pricing a multi-asset option into a
system of linear equations and employ the quantum algorithm due to Harrow,
Hassidim and Lloyd to find its solution. Certain numerical characteristics of
the matrix representing the system of linear equations determine a vital role in
whether computational advantage can be achieved. The central question of this
thesis is whether we can perturb the matrix that is to be inverted in the direction
of more favourable numerical characteristics without compromising the accuracy
of the final solution in representing the present value of the multi-asset option.
Through specific examples, we show that this perturbation does not compromise
the accuracy of the calculated value for the option.
After an introduction to options and their underlying mathematical description,
we provide a derivation for the Black-Scholes equation using stochastic calculus
and its corresponding solution for the vanilla European option through the
Feynman-Kac formula. We continue with the numerical methods of finite difference
approximations to convert the problem into a system of linear equations.
Finally, after presentation of the quantum algorithm, we proceed with numerical
simulations to determine (a) whether the aforementioned perturbation can
be ameliorated with modified boundary conditions and (b) whether a working
end-to-end quantum algorithm for option pricing for the case of a single-asset
European option maybe achieved. Our simulation provides a proof-of-principle
demonstration of the quantum algorithm
Measuring Collective Action Intention Toward Gender Equality Across Cultures
No full textCollective action is a powerful tool for social change and is fundamental to women and girls’ empowerment on a societal level. Collective action towards gender equality could be understood as intentional and conscious civic behaviors focused on social transformation, questioning power relations, and promoting gender equality through collective efforts. Various instruments to measure collective action intentions have been developed, but to our knowledge none of the published measures were subject to invariance testing. We introduce the gender equality collective action intention (GECAI) scale and examine its psychometric isomorphism and measurement invariance, using data from 60 countries ( N = 31,686). Our findings indicate that partial scalar measurement invariance of the GECAI scale permits conditional comparisons of latent mean GECAI scores across countries. Moreover, this metric psychometric isomorphism of the GECAI means we can interpret scores at the country-level (i.e., as a group attribute) conceptually similar to individual attributes. Therefore, our findings add to the growing body of literature on gender based collective action by introducing a methodologically sound tool to measure collective action intentions towards gender equality across cultures
Green Machine Learning: Analysing the Energy Efficiency of Machine Learning Models
No full textThe consumption of energy by Machine Learning
(ML) has increased significantly. There is growing concern about
the sustainable use of ML, where choosing the best ML model
should also consider energy efficiency. The main objective of
the Green Machine Learning paradigm is the simultaneous
optimisation of accuracy and energy consumption. The literature
has presented some suggestions for metrics to be used. However, these metrics have not been extensively compared among
different ML models. To address this aspect, in this paper, we
have analysed six Machine Learning models applied to three
benchmark datasets for binary classification tasks, focusing on
performance and energy consumption. The results of the F1-
Score show that the random forests model outperformed the
other models, while logistic regression was more energy efficient.
These results demonstrate the trade-offs between model performance and energy consumption, providing valuable guidance
for algorithm selection. Performance metrics are an essential
benchmark, with Python’s Scikit-Learn suite of models often
outperforming neural networks in classification tasks. Future
research should extend energy analysis to other machine learning
methods and consider metrics that balance performance and
energy consumption
Optimisation and control of tidal range power plants operation: Is there scope for further improvement?
No full textTidal barrage power plants utilise the tidal range variation to generate clean electricity. Although there are
several operating tidal barrage schemes around the globe, there is still potential to expand the installed capacity.
Given their inherent storage and the high predictability of the tides, tidal barrages can be operated with more
flexibility than many other renewables. This means that the control objective of a barrage operation can vary
from energy maximisation to constant power output, or demand-matching objectives. The operation of a barrage
also influences its impact on the environment and economic activity of the site where it is located, which is a
major cause for the slow deployment of such power plants. The aim of this study is to provide a comprehensive
and critical analysis of the different strategies considered to date to optimise the operation of tidal barrages, with
a focus on an in-depth analysis of the optimisation schemes employed, the barrage models utilised, and op
portunities for further improvement
The Montessori school as a ‘healing’ environment: translating childhood trauma research into effective, trauma-informed, educational practice
No full textBackground: Childhood trauma/adversity is pervasive and has far-reaching consequences for
children’s health and well-being, leading to increased calls for trauma-informed practice (TIP).
Archival data show that early Montessori schools (circa 1907-1917) were recognised as ‘healing’ schools, wherein trauma-affected children improved dramatically.
Aims/objectives: This project aimed to (1) investigate claims of psychological healing in early
Montessori schools; (2) integrate the findings with contemporary knowledge on TIP; (3)
develop a novel Continuing Professional Development (CPD) programme based on this integration; and (4) evaluate its perceived impact on staff in a test school.
Method: A multi-method, three-strand approach was used comprising three distinct and
sequential studies. Study 1 involved a documentary analysis of eyewitness testimonies, media reports, and Montessori’s own accounts of her early schools, to investigate how the Montessori approach supported trauma-affected children. Study 2 integrated the findings of Study 1 with
contemporary trauma literature to develop an innovative CPD programme designed to enhance the capacities of early childhood teachers to support trauma-affected children. Study 3 then
used a case study approach to provide a rich contextual account of teachers’ (n=11) experiences of engaging with this programme, focusing on its perceived impact on their knowledge,
attitudes/beliefs, professional practice, and their views on its feasibility.
Findings: Study 1 identified significant evidence of psychological healing in trauma-affected
children attending Montessori’s early schools. Study 2 found that several features of
Montessori education cohere with contemporary research on TIP approaches, especially the Neurosequential Model in Education (NME), and that these can be integrated to develop a programmme of Montessori-attuned TIP. Study 3 found that early childhood/Montessori teachers rated the new programme highly, stating it positively impacted their practice.
Conclusion: This project makes a significant original contribution to existing knowledge on
Montessori pedagogy and TIP and has important implications for supporting trauma-affected
children in Ireland and elsewhere
Informativeness of the federal reserve chair communication’s sentiment on the monetary policy uncertainty
No full textCan a single personal communication have a significant effect on the uncertainty of the
monetary policy process? We estimate the personal communication risk profile of the U.S.
Federal Reserve (Fed) Chair by using a new dataset of the sentiment revealed by their public
statements during their tenure. We develop a new identification method using the implicit
probability of change of the federal fund rate, and analyze the impact of the Fed communication’s sentiment risk profile on the market price discovery process of interest rates, and
the uncertainty of the monetary policy, in the aftermath of the release of Chair public statements. After controlling for the evolving state of the economy surrounding the meetings,
we find that, based on the heterogeneity across Chairs and their personal traits, there is a
significant statistical and economic difference in the communications’ sentiment, which is
likely to affect the market’s reaction to monetary policy announcements. Specifically, the
sentiment in the Chairs’ communications plays an important role in moderating the potential
surprises in the Fed announcements, and it can be effectively used as a tool for controlling
and measuring monetary policy shocks
Attitudes of millennials toward corporate responsibility: a 28-society multilevel analysis
No full textPurpose
We examined the attitudes of millennial-aged business students toward economic, social and environmental corporate responsibility (CR). Currently, these individuals are of an age that they have entered the workforce and are now ascending or have ascended into roles of leadership in which they have decision-making power that influences their company’s CR agenda and implementation. Thus, following the ecological systems perspective, we tested both the macro influence of cultural values (survival/self-expression and traditional/secular-rational values) and structural forces (income inequality, welfare socialism and environmental vulnerability) on these individuals’ attitudes toward CR.
Design/methodology/approach
This is a multilevel study of 3,572 millennial-aged students from 28 Asian, American, Australasian and European societies. We analyzed the data collected in 2003–2009 using hierarchical linear modeling.
Findings
In our multilevel analyses, we found that survival/self-expression values were negatively related to economic CR and positively related to social CR while traditional/secular-rational values was negatively related to social CR. We also found that welfare socialism was positively related to environmental CR but negatively related to economic CR while environmental vulnerability was not related to any CR. Lastly, income equality was positively related to social CR but not economic or environment responsibilities. In sum, we found that both culture-based and structure-based macro factors, to varying extents, shape the attitudes of millennial-aged students on CR in our sample.
Originality/value
Our study is grounded in the ecological systems theory framework, combined with research on culture, politico-economics and environmental studies. This provides a multidisciplinary perspective for evaluating and investigating the impact that societal (macro-level) factors have on shaping attitudes toward businesses’ engagement in economic, social and environmental responsibility activities. Additionally, our multilevel research design allows for more precise findings compared to a single-level, country-by-country assessment
Walls with no pictures – the alternative interiority of Séamas Mac Annaidh’s short stories
No full textThe Abstract is included in the text