13463 research outputs found
Sort by
A Meta-Learning Approach for Hydrological Time Series Model Selection
Get PDFTime series forecasting is crucial in various fields, with significant socio-economic implications, as accurate predictions can aid in better resource management, disaster preparedness, and economic planning. However, selecting an appropriate forecasting model remains a labor-intensive task demanding expertise. This research introduces a novel meta-learning approach to automate and enhance the model selection process.
We curate extensive time series datasets specific to Ireland, spanning diverse temporal patterns and environmental attributes, including climate data, water level measurements, and landscape characteristics. The initial part of the research focuses on developing a systematic architecture using Extract, Transform, and Load (ETL) technology to integrate heterogeneous data from various sources while ensuring data quality and consistency.
Then, this research concentrates on accurately predicting river water levels. Various Machine Learning (ML) models are employed, relying on previously observed river water level data. The research evaluates the predictive performance of these ML models across all hydrometric stations in Ireland and demonstrates the importance of careful model selection based on geographic and hydrological features. The results demonstrated that a universal ‘one-model-fits-all’ approach is not suitable for hydrological time series data.
Subsequently, this research explores the core contribution of applying meta-learning to context-aware model selection for river water-level prediction. The study demonstrates that meta-learning enhances the accuracy and reliability of hydrologic time series forecasting, addressing the complexities of this task and providing valuable insights into applying ML in this domain. The efficacy of our meta-learning approach is evaluated across various real-world time series datasets, consistently demonstrating its superiority over traditional model selection techniques. Importantly, our approach streamlines and expedites time series forecasting, making it more accessible to researchers.
In conclusion, this thesis significantly contributes to ML-based environmental time-series data prediction using a model-selection meta-learner approach and enhanced data integration techniques. The results show that our research aligns with the growing trend of automated machine learning and has the potential to revolutionise time series forecasting in diverse applications
Evaluating and Mitigating Transaction Costs with Recurrent Neural Networks
Get PDFThis thesis develops a method for evaluating and mitigating the effect of transaction costs on trading strategies with many assets. An iteration procedure yields the cost-adjusted portfolio return, enabling the formulation of portfolio-choice problems as optimization of Recurrent Neural Networks (RNN). This method reproduces the theoretical results available for one risky asset and the numerical approximations available
for two risky assets through finite-elements. Crucially, the RNN model scales to several assets and is fully interpretable, as its parameters identify their no-trade region. Importance-sampling significantly enhances the model’s performance, especially with several assets. An application to equally-weighted funds demonstrates the method’s ability to reduce both tracking error and tracking difference from an empirical target
Analysis of the emerging contribution of the Spanish insurance sector to the development of sustainable finance
No full textThe purpose of our investigation is to analyse the contribution of the insurance industry to sustainable finance in Spain and how this, in turn, is transforming the insurance industry itself. From there, we will also explore how insurers themselves and the concept of sustainable finance have an impact on Sustainable Development Goals (SDG) and how they may help to accomplish them. The proposed research methodology includes a section based on studies conducted by the sector’s employers’ association in Spain and another based on a series of interviews with experts and reference specialists. This is combined with an evaluation of best practices from leading entities, as well as a bibliographical review from institutional sources and monographs. Using these methods, we have demonstrated that insurance companies add value to sustainable finance from a triple perspective: as investors with high resource mobilisation potential; as expert risk managers, which now include
environmental, social, and governance risks (ESG); and through the insurance business itself, through subscription policies and product and services creation/modification. Moving forward, we identified several areas of work which will be critical in defining the full potential of the insurance sector in the development of sustainable finance, such as EU regulatory development, integration of ESG risks into internal models, access to sustainable investment offer, and stakeholder valuation
Synthetic Time Series for Anomaly Detection in Cloud Microservices
Get PDFThis paper proposes a framework for time series generation built to investigate anomaly detection in cloud microservices. In the field of cloud computing, ensuring the reliability of microservices is of paramount concern and yet a remarkably challenging task. Despite the
large amount of research in this area, validation of anomaly detection algorithms in realistic environments is difficult to achieve. To address this challenge, we propose a framework to mimic the complex time series patterns representative of both normal and anomalous cloud microservices behaviors.We detail the pipeline implementation that allows deployment and management of microservices as well as the theoretical approach required to generate anomalies. Two datasets generated using the proposed framework have been made publicly available through GitHub
Secure and Decentralized Collaboration in Oncology: A Blockchain Approach to Tumor Segmentation
Get PDFThis research presents an innovative framework that uses blockchain technology to improve tumor segmentation in medical imaging. The approach tackles issues related to data security, particularly when dealing with real private dataset, annotation accuracy, and collaboration. With the growing reliance of the medical industry on accurate tumor segmentation from medical images for cancer diagnosis and treatment, current methods are inadequate in maintaining data accuracy and promoting collaboration among experts across different countries. Our suggested approach utilizes blockchain technology to establish a decentralized, secure platform for the collaborative obtaining, annotation, and validation of medical
images by data scientists, oncologists, and radiologists. Smart
contracts streamline essential procedures such as verification of
annotations, consensus among experts, and remuneration of
contributors, guaranteeing the dependability and excellence of
the data. Furthermore, the unchangeable record of transactions
in the blockchain ensures a reliable basis for implementing
artificial intelligence and machine learning algorithms. This
improves the accuracy of segmenting data and allows for
predictive modeling. This strategy not only improves the
precision and effectiveness of tumor segmentation but also
promotes a worldwide collaborative environment, which has the
potential to revolutionize cancer diagnostics and treatment
planning. Furthermore, it ensures the privacy and security of
patient data
One-dimensional topological channels in heterostrained bilayer graphene
Get PDFThe domain walls between AB- and BA-stacked gapped bilayer graphene have garnered intense interest as they host topologically protected, valley-polarized transport channels. The introduction of a twist angle between the bilayers and the associated formation of a moiré pattern has been the dominant method used to study these topological channels, but heterostrain can also give rise to similar stacking domains and interfaces. Here, we theoretically study the electronic structure of a uniaxially heterostrained bilayer graphene. We discuss the formation and evolution of interface-localized channels in the one-dimensional moiré pattern that emerges due to the different stacking registries between the two layers. We find that a uniform heterostrain is not sufficient to create one-dimensional topological channels in biased bilayer graphene. Instead, using a simple model to account for the in-plane atomic reconstruction driven by the changing stacking registry, we show that the resulting expanded Bernal-stacked domains and sharper interfaces are required for robust topological interfaces to emerge. These states are highly localized in the AA- or SP-stacked interface regions and exhibit differences in their layer and sublattice distribution depending on the interface stacking. We conclude that heterostrain can be used as a mechanism to tune the presence and distribution of topological channels in gapped bilayer graphene systems, complementary to the field of twistronics
The politics of Ireland’s food supply, 1895 to 1923: through peace, war, revolution, and partition
No full textDespite the centrality of the Great Famine in Irish history, the politics of food supply during the late nineteenth and early twentieth centuries has received little historical attention, unlike Britain, Europe, and the USA. This dissertation assesses Ireland’s food supply politics in peacetime between 1895 and 1914 and during wartime conditions until 1923. Quantitative analysis of poor law relief data creates the first proxy measure of late nineteenth-century food supply levels. The measure establishes that contrary to the historiographical view, the 1898 food supply crisis was the most severe of three crises in the 1890s.
A qualitative exploration of the 1898 crisis delivers the first comprehensive synthesis of the politics of food shortage since the Great Famine. It reveals the deeply politicised interplay of the cooperative movement, government, parliamentarians, clergy, nuns, shopkeepers, English women, Irish activists, newspapers and photography. Uniquely, this research identifies how William O’Brien and Michael Davitt successfully exploited the 1898 food crisis for their early mobilisation of the United Irish League. It also examines two food supply crises in the 1900s, school meals activism and Roger Casement’s funding of school meals in Connemara.
This research demonstrates that Britain pursued two equally important strategic needs in Ireland during the First World War — military recruitment and food exports. It creates the first history of how the government gradually but firmly sequestered control of Ireland’s food production and supply away from the free market. It also uncovers how, a year before the Dáil Éireann counter state, the Sinn Féin Food Department operated nationwide in early 1918 as an executive — governmental — authority, purchasing, storing and distributing food supplies. Analysis of food supply dynamics provides new insights into the complex transition from constitutional to advanced nationalist politics and the early contrasting food supply fortunes of Northern Ireland and the Irish Free State
Secure and Decentralized Collaboration in Oncology: A Blockchain Approach to Tumor Segmentation
Get PDFThis research presents an innovative framework that uses blockchain technology to improve tumor segmentation in medical imaging. The approach tackles issues related to data security, particularly when dealing with real private dataset, annotation accuracy, and collaboration. With the growing reliance of the medical industry on accurate tumor segmentation from medical images for cancer diagnosis and treatment, current methods are inadequate in maintaining data accuracy and promoting collaboration among experts across different countries. Our suggested approach utilizes blockchain technology to establish a decentralized, secure platform for the collaborative obtaining, annotation, and validation of medical images by data scientists, oncologists, and radiologists. Smart contracts streamline essential procedures such as verification of annotations, consensus among experts, and remuneration of contributors, guaranteeing the dependability and excellence of the data. Furthermore, the unchangeable record of transactions in the blockchain ensures a reliable basis for implementing artificial intelligence and machine learning algorithms. This improves the accuracy of segmenting data and allows for predictive modeling. This strategy not only improves the precision and effectiveness of tumor segmentation but also promotes a worldwide collaborative environment, which has the potential to revolutionize cancer diagnostics and treatment planning. Furthermore, it ensures the privacy and security of patient data
VidBasys: A User-friendly Interactive Video Retrieval System for Novice Users in IVR4B
Get PDFIn this paper, we present the VidBasys interactive video retrieval system for novice users, an upgraded version of VideoCLIP 2.0 that participated in the Video Browser Showdown 2024. While the novel user interface is designed in a more user-friendly way for newbies to accommodate the target of the Interactive Video Retrieval for Beginner (IVR4B), the core search engine is enhanced with the advance of the recent CLIP model to bridge the gap in semantics between image and text. This version is designed to focus on novice users with a simple, easyto-use but effective user interface. The system supports freetext search to enhance the user experience and minimise the number of actions required for filtering. The new user interface supports simple search and filters with clearly designed freetext search boxes. In addition, the retrieved results are displayed in an optimised layout to maximise image display space and minimise user interactions. The improvements are expected to support novice users in accurately retrieving the desired video
A phenomenological exploration of the lived experience of adults experiencing pathological demand avoidance
Get PDFThis study explores the lived experiences of adults with Pathological Demand Avoidance (PDA) in Ireland. There is a paucity of research exploring the experience of those living with PDA in Ireland which impacts levels of awareness and understanding of anxiety-based demand avoidance and its intersection with autism. As lifelong conditions, this has implications for adults who can struggle to access services or appropriate supports or can be isolated in the transition from Child and Adolescent services. This paper extracts data from a national study, Mapping Experiences of Pathological Demand Avoidance in Ireland, and examines the lived experience and personal histories of adults with a diagnosis of / self-identifying with PDA and autism, consisting of statements extracted from open-ended survey responses, and in-depth interviews. A deductive approach to the analysis of their testimonies discovered four superordinate themes: (1) Bidirectional social challenges, (2) Life experiences, (3) Trust and safety and (4) Accepting our truth. Subordinate leitmotifs describe: Pervasive anxiety, Challenges negotiating life demands and Flexibility in education settings. This paper reveals the interplay between PDA, autism and mental health for these participants combined, leading to significant challenges in daily life. To support better life quality and flourishing, participants advised greater autonomy and flexibility of support across all aspects of life and more awareness of PDA across society