114612 research outputs found
Sort by
In search of effective corporate grievance mechanisms : can mandatory due diligence laws be a progressive force?
No full text
A linear transportation Lp distance for pattern recognition
Get PDFThe transportation Lp distance, denoted TLp, has been proposed as a generalisation of Wasserstein Wp distances motivated by the property that it can be applied directly to colour or multi-channelled images, as well as multivariate time-series without normalisation or mass constraints. These distances, as with Wp, are powerful tools in modelling data with spatial or temporal perturbations. However, their computational cost can make them infeasible to apply to even moderate pattern recognition tasks. We propose linear versions of these distances and show that the linear TLp distance significantly improves over the linear Wp distance on signal processing tasks, whilst being several orders of magnitude faster to compute than the TLp distance
Deciding atomicity of subword-closed languages
No full textWe study languages closed under the non-contiguous (scattered) subword containment order. Any subword-closed language L can be uniquely described by its anti-dictionary, i.e. the set of minimal words that do not belong to L. For a language over a finite alphabet, the anti-dictionary is necessarily finite. A language L is said to be atomic if it cannot be presented as the union of two subword-closed languages different from L. In this work, we provide a decision procedure which, given a language over a finite alphabet defined by its anti-dictionary, decides whether it is atomic or not. We also develop an algorithmic procedure for decomposing a language, which is not atomic, into finitely many atomic sublanguages
Using symbiotic empirical ethics to explore the significance of relationships to clinical ethics : findings from the Reset Ethics research project
No full textAt the beginning of the coronavirus (Covid-19) pandemic, many non-Covid healthcare services were suspended. In April 2020, the Department of Health in England mandated that non-Covid services should resume, alongside the continuing pandemic response. This ‘resetting’ of healthcare services created a unique context in which it became critical to consider how ethical considerations did (and should) underpin decisions about integrating infection control measures into routine healthcare practices. We draw on data collected as part of the ‘NHS Reset Ethics’ project, which explored the everyday ethical challenges of resetting England’s NHS maternity and paediatrics services during the pandemic.
Methods
Healthcare professionals and members of the public participated in interviews and focus group discussions. The qualitative methods are reported in detail elsewhere. The focus of this article is our use of Frith’s symbiotic empirical ethics methodology to work from our empirical findings towards the normative suggestion that clinical ethics should explicitly attend to the importance of relationships in clinical practice. This methodology uses a five-step approach to refine and develop ethical theory based on a naturalist account of ethics that sees practice and theory as symbiotically related.
Results
The Reset project data showed that changed working practices caused ethical challenges for healthcare professionals, and that infection prevention and control measures represented harmful barriers to the experience of receiving and offering care. For healthcare professionals, offering care as part of a relational interaction was an ethically important dimension of healthcare delivery.
Conclusions
Our findings suggest that foregrounding the importance of relationships across a hospital community will better promote the ethically important multi-directional expression of caring between healthcare professionals, patients, and their families. We offer two suggestions for making progress towards such a relational approach. First, that there is a change of emphasis in clinical ethics practice to explicitly acknowledge the importance of the relationships (including with their healthcare team) within which the patient is held. Second, that organisational decision-making should take into account the moral significance afforded to caring relationships by healthcare professionals, and the role such relationships can play in the negotiation of ethical challenges
1973 and the American horror film : political futurity in The Exorcist and The Texas Chain Saw Massacre
No full textThis article argues that two classic films made in 1973, William Friedkin’s The Exorcist and Tobe Hooper’s The Texas Chain Saw Massacre, both anticipate and give nightmarish form to an underlying political shift that historians often originate at exactly the same moment: the advent of the neo-liberal/neo-conservative order that would come to define American life for at least the next 40 years. Rather than seeing these films as centred on ancient demons or obsolete workers, this essay reverses the standard ‘Gothic temporality’ of the past’s persistence and positions the horror film as, instead, a form of speculative fiction; not a registration of history’s traumatic aftermath but a barometer of the emerging political future
Non-negative matrix factorization for link prediction preserving row and column spaces
No full textNon-negative Matrix Factorization (NMF) has been widely adopted for link prediction, aiming at finding multiple low-dimensional matrices whose product approximates the adjacency matrix of a network. Most existing NMF-based models incorporate auxiliary information with well-defined geometric meanings, but there is no evidence that they have reasonable mathematical interpretations. In this paper, we propose a model, NMF-CR, that incorporates both row-space and column-space information into the NMF framework. NMF-CR not only carries well-defined geometric meanings but also boasts a reasonable mathematical interpretation. Moreover, we provide efficient updating rules to infer the parameters of NMF-CR with guaranteed convergence. Extensive experiments demonstrate that our model achieves higher prediction accuracy than its competitors
Future research priorities for soft-tissue knee injuries
No full textTo identify unanswered questions about the prevention, diagnosis, treatment, and rehabilitation and delivery of care of first-time soft-tissue knee injuries (ligament injuries, patella dislocations, meniscal injuries, and articular cartilage) in children (aged 12 years and older) and adults. The James Lind Alliance (JLA) methodology for Priority Setting Partnerships was followed. An initial survey invited patients and healthcare professionals from the UK to submit any uncertainties regarding soft-tissue knee injury prevention, diagnosis, treatment, and rehabilitation and delivery of care. Over 1,000 questions were received. From these, 74 questions (identifying common concerns) were formulated and checked against the best available evidence. An interim survey was then conducted and 27 questions were taken forward to the final workshop, held in January 2023, where they were discussed, ranked, and scored in multiple rounds of prioritization. This was conducted by healthcare professionals, patients, and carers. The top ten included questions regarding prevention, diagnosis, treatment, and rehabilitation. The number one question was, 'How urgently do soft-tissue knee injuries need to be treated for the best outcome?'. This reflects the concerns of patients, carers, and the wider multidisciplinary team. This validated process has generated ten important priorities for future soft-tissue knee injury research. These have been submitted to the National Institute for Health and Care Research. All 27 questions in the final workshop have been published on the JLA website. [Abstract copyright: © 2024 The British Editorial Society of Bone & Joint Surgery.
To each (textual sequence) its own : improving memorized-data unlearning in large language models
No full textLLMs have been found to memorize training textual sequences and regurgitate verbatim said sequences during text generation time. This fact is known to be the cause of privacy and related (e.g., copyright) problems. Unlearning in LLMs then takes the form of devising new algorithms that will properly deal with these side-effects of memorized data, while not hurting the model's utility. We offer a fresh perspective towards this goal, namely, that each textual sequence to be forgotten should be treated differently when being unlearned based on its degree of memorization within the LLM. We contribute a new metric for measuring unlearning quality, an adversarial attack showing that SOTA algorithms lacking this perspective fail for privacy, and two new unlearning methods based on Gradient Ascent and Task Arithmetic, respectively. A comprehensive performance evaluation across an extensive suite of NLP tasks then mapped the solution space, identifying the best solutions under different scales in model capacities and forget set sizes and quantified the gains of the new approaches
Modeling and analysis for material removal and surface roughness in fluid jet polishing of optical glass
No full textFluid jet polishing (FJP) is a non-contact polishing technology that can fabricate free-form optical surfaces with sub-micron-level form accuracy and nano-level surface roughness, especially for hard and brittle materials. The surface generation model of FJP can be used to guide the determination and optimization of process parameters and is of great significance for understanding the evolution mechanism of surface microtopography. However, predictive models for the microscopic topography of polished surfaces are still lacking. This study established a macroscopic surface profile model for predicting 3D material removal characteristics and surface texture by combining the 3D computer fluid dynamics (CFD) simulation model and single-particle erosion mechanism. A fractal theory-based erosion model has been built to calculate the material removal caused by the erosion of a single abrasive particle on the rough surface; thus, it predicts the micro-topography and surface roughness of the polished samples. A series of polishing experiments were conducted to analyze the feasibility and accuracy of the model quantitatively and study the influence mechanism of process parameters on the material removal characteristics and surface quality. Results indicated that the models could well predict material removal and surface roughness. The prediction accuracy of the surface roughness Ra and maximum removal depth is better than 91.6% and 90%, respectively. It is also found that the material removal rate of FJP could reach 0.517 mm3/min, and the surface roughness convergence rate could reach 62.9%