Repository of the University of Namur
Not a member yet
94692 research outputs found
Sort by
Spatial interpolation of health and demographic variables:Predicting malaria indicators with and without covariates
No full textAccurate mapping and disaggregation of key health and demographic risk factors have become increasingly important for disease surveillance, which can reveal geographical social inequalities for improved health interventions and for monitoring progress on relevant Sustainable Development Goals (SDGs). Household surveys like the Demographic and Health Surveys have been widely used as a proxy for mapping SDG-related household characteristics. However, there is no consensus on the workflow to be used, and different methods have been implemented with varying complexities. This study aims to compare multiple modelling frameworks to model indicators of human vulnerability to malaria (SDG Target 3.3) in Senegal. These indicators were categorised into socioeconomic (e.g., stunting prevalence, wealth index) and malaria prevention indicators (e.g., indoor residual spraying, insecticide-treated net ownership). We compared three categories of the commonly used methods: (1) spatial interpolation methods (i.e., inverse distance weighting, thin plate splines, kriging), (2) ensemble methods (i.e., random forest), and (3) Bayesian geostatistical models. Most indicators could be modelled with medium to high predictive accuracy, with R2 values ranging from 0.40 to 0.86. No method or method category emerged as the best, but performance varied widely. Overall, socioeconomic indicators were generally better predicted by covariate-based models (e.g., random forest and Bayesian models), while methods using spatial autocorrelation alone (e.g., thin plate splines) performed better for variables with heterogeneous spatial structure, such as ethnicity and malaria prevention indicators. Increasing the complexity of the models did not always improve predictive performance, e.g., thin plate splines sometimes outperformed random forest or Bayesian geostatistical models. Beyond performance, we compared the different methods using other criteria (e.g., the ability to constrain the prediction range or to quantify prediction uncertainty) and discussed their implications for selecting a modelling approach tailored to the needs of the end user.</p
Nouveaux regards sur l’avouerie:Les avoués des abbayes et des sièges épiscopaux entre Loire et Rhin (fin IXe – mil. XIIIe siècle)
No full textTHE FEMININE ARCHETYPES IN ROSA MONTERO’S LÁGRIMAS EN LA LLUVIA:MODALITIES AND FUNCTIONS
Get PDFThis paper examines the female archetypes that in Lágrimas en la lluvia, a novel published by Rosa Montero in 2011, are drawn from both canonical literature and Western pop culture —from the femme fatale to the domestic angel, passing through Lara Croft or the prototypical chicklit protagonist— to adapt them to the molds of dystopian science fiction. After showing to what extent the construction of the protagonist of the story, Bruna Husky -in which most of these archetypes converge- is based on a certain generic syncretism (in the double sense, sexual and architextual), we will analyse the formal (narrative and hypertextual) and thematic (through topoï and stereotyping phenomena) devices for the construction of female archetypes that are organised around two functions traditionally associated with the representation of women in the Western imaginary: that of the active sensual subject and that of the passive object, focused from the sphere of the intimate, linked in turn to the concept of care. From there we will show how the articulation of these figures with a fairly traditional science-fiction encyclopaedia (populated by cyborgs, artificial planets, plasma weapons and aliens), as well as the hybridism between detective novel, romance novel and bildungsroman, participate in the paratopia (in Maingueneau’s terms) of Bruna’s character, while enabling an ambivalent poetics of stereotyping that appeals to the reader’s expectations in a playful reading pact.</p
Mixed Molybdenum–Tungsten Oxide as Dual-Band, VIS–NIR Selective Electrochromic Material
No full textThe energy efficiency of smart windows can be greatly improved by integrating dual-band electrochromic materials based on nanostructured doped metal oxides, as these allow for the dynamic and independent control of light and heat supplies in buildings, respectively, related to transmitted visible (VIS) and near-infrared (NIR) solar radiations. Mixed molybdenum–tungsten oxides MoyW1−yO3−δ “MoWOx” are considered as innovative compounds in this context: in comparison with parent WO3−δ formulations, a remarkable increase of unpaired electrons can be obtained from the formation of a larger amount of both reduced species and oxygen vacancies during the synthetic process of MoWOx materials, with a consequently boosted intensity of light absorption by 850–900 nm wavelengths. Consecutively, spin-coated films obtained from MoWOx structures, solvothermally synthesized for 1 h from 2/1 Mo/W atomic ratios, are shown to be electrochromically commutable through a noticeable VIS-darkened and NIR-transparent warm state, while WO3−δ benchmarks classically evolve between bright (VIS- and NIR-bleached), cool (VIS-bleached, NIR-opaque), and dark (VIS- and NIR-opaque) states. All in all, such advanced optical functionalities are of high interest for fine-tuning the selectivity and improving the efficiency of electrochromic fenestration solutions, further improving their capacity to adapt to different climates, seasons, and user preferences.</p
Preventing rather than ‘curing’ artificial intelligence-induced discrimination:Article 10 of the AI Act and the prevention of bias in high-risk artificial intelligence systems through data governance
No full textDans la doctrine juridique, la lutte contre les discriminations induites par l’intelligence artificielle a principalement été envisagée sous l’angle du droit de la non-discrimination et du droit de la protection des données. Ces moyens de droit sont principalement actionnés a posteriori. Depuis l’adoption du règlement sur l’intelligence artificielle (AI Act), l’accent est désormais mis sur la prévention. L’un des objectifs du règlement est en effet de compléter le droit de l’Union existant en matière de non-discrimination en réduisant le risque de discrimination induite par l’IA. Ce complément est constitué à titre principal de l’article 10 de l’AI Act, lequel contient des mesures concernant la conception et la qualité des jeux de données utilisés pour le développement de systèmes d’IA à haut risque ou plus largement la « gouvernance des données ». La disposition n’échappe cependant pas à l’une des critiques principales adressées au règlement : son caractère ésotérique. L’imprécision de certains termes, leur technicité, ainsi que la mobilisation de différentes disciplines et de divers domaines du droit rendent la disposition difficile à interpréter. La présente contribution propose donc de clarifier la portée de cet article, tout en interrogeant la pertinence des mesures adoptées aux fins d’atteindre son but : réduire la survenance des discriminations induites par l’IA.In legal doctrine, the fight against discrimination induced by artificial intelligence has primarily been approached from the perspective of non-discrimination law and data protection law. These legal mechanisms are generally applied retrospectively. Since the adoption of the AI Act, the focus has shifted toward prevention. One of the regulation's objectives is indeed to complement existing Union law on non-discrimination by reducing the risk of AI-induced discrimination. This complement primarily takes the form of Article 10 of the AI Act, which sets out measures concerning the design and quality of datasets used in the development of high-risk AI systems—or more broadly, data governance. However, this provision is not exempt from one of the main criticisms leveled at the regulation: its esoteric nature. The vagueness of certain terms, their technical complexity, and the intersection of various disciplines and areas of law make the provision difficult to interpret. This contribution therefore aims to clarify the scope of Article 10 while questioning the relevance of the adopted measures in achieving its objective: reducing the occurrence of AI-induced discrimination
