Irish Universities
Not a member yet
78452 research outputs found
Sort by
Good Counterfactuals and Where to Find Them: A Case-Based Technique for Generating Counterfactuals for Explainable AI (XAI)
No full textThe 28th International Conference on Case-Based Reasoning (ICCBR 2020), Salamanca, Spain, 8–12 June 2020 (held online due to COVID-19 pandemic)Recently, a groundswell of research has identified the use of counter-factual explanations as a potentially significant solution to the Explainable AI (XAI) problem. It is argued that (i) technically, these counterfactual cases can be generated by permuting problem-features until a class-change is found, (ii) psychologically, they are much more causally informative than factual explanations, (iii) legally, they are GDPR-compliant. However, there are issues around the finding of “good” counterfactuals using current techniques (e.g.sparsity and plausibility). We show that many commonly-used datasets appear to have few “good” counterfactuals for explanation purposes. We propose a new case-based approach for generating counterfactuals, using novel ideas about the counterfactual potential and explanatory coverage of a case-base. The new technique reuses patterns of good counterfactuals, present in a case-base, to generate analogous counterfactuals that can explain new problems and their solutions. Several experiments show how this technique can improve the counterfactual potential and explanatory coverage of case-bases, that were previously found wanting.Science Foundation IrelandInsight Research Centr
REMOS-IoT - A Relay and Mobility Scheme for Improved IoT Communication Performance
No full textThe Internet of Things (IoT) can avail from device-to-device (D2D) communication techniques to increase object data exchange performance. IoT networks aim to offer a massive number of services at high quality levels, and many of the devices providing these services are mobile. Devices such as wearables, sensors, drones and smart vehicles need constant connectivity despite their moving patterns and therefore, an IoT architecture should consider both Quality of Service (QoS) and mobility. D2D allows devices to communicate directly to share content and functionality, such as access to the Internet. This paper proposes REMOS-IoT - A RElay and MObility Scheme for improved IoT communication performance in support of increased QoS for the data exchange services between mobile IoT devices. Simulation-based testing showed how performance of devices increased in several scenarios, demonstrating the efficiency of the proposed architecture and algorithms
Staff member\u27s views and attitudes to supporting people with an Intellectual Disability: A multi-method investigation of intimate relationships and sexuality
No full textBackground: Staff member\u27s views and attitudes can have a significant impact on people with an intellectual disability. This study explores staff attitudes and experiences of people with an intellectual disability\u27s expression of relationships and sexuality across service providers in Ireland. Methods: Staff (n = 86) from service providers (n = 5) completed the Attitudes to Sexuality Questionnaire-Individuals with an Intellectual Disability (ASQ-ID) and open-ended questions. Results: Multidirectional significant differences were noted on staff attitudes to sexuality based on demographic factors, that is, age, education, as well as non demographics, for example, attendance at training, and active practising of religion. Qualitatively, three themes emerged: “Unsupported and Frustration”; “Taboo Subject Matter” and “Vulnerability and Access to Education”. Staff reported a lack of service supports, dominant conservative cultures and people with an intellectual disability\u27s vulnerability. Conclusion: This study highlights the need for improved service practices and enhanced provisions for staff and the people with an intellectual disability that they support
Bayesian Case-Exclusion and Explainable AI (XAI) for Sustainable Farming
No full textThe 29th International Joint Conference on Artificial Intelligence - 17th Pacific Rim International Conference on Artificial Intelligence (IJCAI-PRICAI-20), Yokohama, Japan, January 2021 (Conference postponed due to COVID-19 pandemic)Smart agriculture (SmartAg) has emerged as a rich domain for AI-driven decision support systems (DSS); however, it is often challenged by user-adoption issues. This paper reports a case-based reasoning system, PBI-CBR, that predicts grass growth for dairy farmers, that combines predictive accuracy and explanations to improve user adoption. PBI-CBR’s key novelty is its use of Bayesian methods for case-base maintenance in a regression domain. Experiments report the tradeoff between predictive accuracy and explanatory capability for different variants of PBI-CBR, and how updating Bayesian priors each year improves performance.Science Foundation IrelandInsight Research Centr
General mechanisms for non-additivity between multiple stressors at higher temporal and biological scales
No full textMultiple anthropogenic stressors threaten the diversity, stability and functioning of ecosystems worldwide. The combined effects of these stressors can, however, be surprisingly difficult to predict due to complex interactions across different levels of organisation. Ecosystem managers are particularly concerned by synergistic stressor interactions, which cause the combined effects of stressors to be greater than expected based on their individual effects alone. Multiple-stressor researchers have primarily used phenomenological approaches to search for generalities in the frequencies of stressor interactions, but have so far encountered an overwhelming amount of context-dependence. Consequently, there is growing interest in a mechanistic understanding of the interactions between stressors (Chapter 2). In this thesis, I integrate evolutionary and ecological theory into the empirical field of multiple-stressor research to describe mechanisms of non-additivity between stressors, that are not specific to particular species, ecosystems or stressors. I focus on general sources of non-additivity at higher temporal scales (Chapter 3) and at higher levels of biological organisation (Chapters 4 and 5). An overarching theme of this thesis is that non-additivity between stressors can not only be caused by natural effects, but can also arise due to observational effects. Natural sources of non-additivity can be physico-chemical interactions between the stressors themselves, or biological effects occurring at multiple levels of organisation from individuals to ecosystems. Conversely, observational sources of non-additivity stem from statistical effects during the measurement and prediction of responses to stressors. These effects can obscure the natural sources of non-additivity that researchers are actually interested in. Recognising these contrasting sources of non-additivity is an important step towards the accurate prediction of the ecological impacts of multiple stressors. A mechanistic understanding of stressor interactions will enhance the management and conservation of ecosystems in the Anthropocene
Ireland in the Reign of Henry VIII: The Making of Tudor Political Theology, 1515-47
No full textThe reign of Henry VIII was a watershed in Irish history. Historians, however, have underestimated the impact of the Henrician Reformation. Exploring the making of Tudor political theology against the North Atlantic and Eurasian arenas of inter-imperial, dynastic, and spiritual intrigue and strife in which struggle for sovereignty in Ireland unfolded, this thesis argues that the break with Rome coalesced with the ramifications of the Kildare Rebellion (1534-5) and the Act of Kingly Title (1541) to alter the terms of Christian sovereignty and moral governance in Ireland. As the turmoil of the 1530s-40s convulsed the tenets of divine and civil order, a distinctly Reformation political theology of ‘truth’ rooted in Henrician theocratic and sacramental kingship emerged. The consequences for the terms of Tudor ‘reform’ and Christian crown subject-hood and service were steep. For what obedience to God and king entailed, what ‘civility’, conformity, and ‘sincere’ submissions to the crown connoted, and how these aligned with ‘true crown service’ changed as the perceived usurpation of princely imperium in Ireland – and the related instabilities of the viceroyal office – acquired a new urgency, ongoing international intrigue and factional strife threatened Tudor sovereignty, and the status of the sacred became violently contested. In this volatile environment further plagued by mounting accusations of treason and the proliferation of oaths to secure rival bonds of spiritual and civil allegiance to foreign powers, ‘religion’ in Ireland, encompassing far more than mere devotional life, was not simply ‘Catholicism without the pope’ , but was transformed into a distinctly Anglo-Irish variant of post-Reformation Henrician Christianity
If sex were a factor… The Securitate Archives and issues of morality in documents related to religious life
No full textThe issue of morality is a vector in the analysis of the archival documents related to religion in communism. When the veil of privacy is lifted and the secret is no more, a rich picture unfolds for the researcher. Blackened names, the minimal protection offered to the actors that surface in surveillance files will do little in affording the subject of such files the privacy his/her actions were thought to have been acted in. For clergy and church members alike the moral stick they are measured against is higher than for the rest. It was self-imposed in many cases. Documents of the CNSAS archives on religion abound with stories about sexuality in many forms. Judges of the morality of the life of “God’s men” the Securitate officers will highlight the failures of the clergy based on guidelines that pertain to the church rather than the Securitate. This article is an overview of the way in which morality permeated the Securitate documents on religious life in communist Romania. How it was used and the reasons behind it. It answers questions related to the disappearance of the morality standard in other archives that dealt specifically with religion in communism
Big (pig) data and the internet of the swine things: a new paradigm in the industry
No full textImplications
• Big data collected on farms can be transformed into useful
information to improve decision making and maximize productivity. A swine management system consisting of tools
(software and devices), with a protocol and standard operative procedures, can generate the necessary information for
the decision-making process.
• New technologies such as electronic feeders and artificial intelligence systems capturing big data will provide a better understanding of animal requirements and behavior, increasing
efficiency and sustainability.
• Biosecurity can be improved using tracking devices for
farm staff, recording movements real-time to decrease disease risks and consequently, improve health and productive
performance
Extraction of Protein from Four Different Seaweeds Using Three Different Physical Pre-Treatment Strategies
No full textSeaweeds are a rich source of protein and can contain up to 47% on the dry weight basis. It is challenging to extract proteins from the raw biomass of seaweed due to resilient cell-wall complexes. Four species of macroalgae were used in this study-two brown, Fucus vesiculosus and Alaria esculenta, and two red, Palmaria palmata and Chondrus crispus. Three treatments were applied individually to the macroalgal species: (I) high-pressure processing (HPP); (II) laboratory autoclave processing and (III) a classical sonication and salting out method. The protein, ash and lipid contents of the resulting extracts were estimated. Yields of protein recovered ranged from 3.2% for Fucus vesiculosus pre-treated with high pressure processing to 28.9% protein recovered for Chondrus crispus treated with the classical method. The yields of protein recovered using the classical, HPP and autoclave pre-treatments applied to Fucus vesiculosus were 35.1, 23.7% and 24.3%, respectively; yields from Alaria esculenta were 18.2%, 15.0% and 17.1% respectively; yields from Palmaria palmata were 12.5%, 14.9% and 21.5% respectively, and finally, yields from Chondrus crispus were 35.2%, 16.1% and 21.9%, respectively. These results demonstrate that while macroalgal proteins may be extracted using either physical or enzymatic methods, the specific extraction procedure should be tailored to individual species
High-accuracy facial depth models derived from 3D synthetic data
No full textIn this paper, we explore how synthetically generated 3D face models can be used to construct a high-accuracy ground truth for depth. This allows us to train the Convolutional Neural Networks (CNN) to solve facial depth estimation problems. These models provide sophisticated controls over image variations including pose, illumination, facial expressions and camera position. 2D training samples can be rendered from these models, typically in RGB format, together with depth information. Using synthetic facial animations, a dynamic facial expression or facial action data can be rendered for a sequence of image frames together with ground truth depth and additional metadata such as head pose, light direction, etc. The synthetic data is used to train a CNN-based facial depth estimation system which is validated on both synthetic and real images. Potential fields of application include 3D reconstruction, driver monitoring systems, robotic vision systems, and advanced scene understanding