179566 research outputs found
Sort by
Designing Transport-Level Encryption for Datacenter Networks
Get PDFCloud applications need network data encryption to isolate from other tenants and protect their data from potential eavesdroppers in the network infrastructure. This paper presents SDT, a protocol design for emerging datacentertransport protocols, such as NDP and Homa, to integrate data encryption. SDT uses per-message record sequence number spaces in a secure session, which ensures unique message identities for its messages to prevent replay attacks. This design enables transport-level encryption that supports existing NIC offloads designed for TLS over TCP, native protocol number alongside TCP and UDP, and message-based abstraction that mitigates head-of-line blocking and enables the network or host stack to identify the message boundaries for load balancing. We implement SDT in the Linux kernel by extending Homa/Linux and improves RPC throughput by up to 41 % and latency by up to 35 % in comparison to TLS/TCP
Learn to Rank Risky Investors: A Case Study of Predicting Retail Traders’ Behaviour and Profitability
Get PDFIdentifying risky traders with high profits in financial markets is crucial for market makers, such as trading exchanges, to ensure effective risk management through real-time decisions on regulation compliance and hedging. However, capturing the complex and dynamic behaviours of individual traders poses significant challenges. Traditional classification and anomaly detection methods often establish a fixed risk boundary, failing to account for this complexity and dynamism. To tackle this issue, we propose a profit-aware risk ranker (PA-RiskRanker) that reframes the problem of identifying risky traders as a ranking task using Learning-to-Rank (LETOR) algorithms. Our approach features a Profit-Aware binary cross entropy (PA-BCE) loss function and a transformer-based ranker enhanced with a self-cross-trader attention pipeline. These components effectively integrate profit and loss (P&L) considerations into the training process while capturing intra- and inter-trader relationships. Our research critically examines the limitations of existing deep learning-based LETOR algorithms in trading risk management, which often overlook the importance of P&L in financial scenarios. By prioritising P&L, our method improves risky trader identification, achieving an 8.4 score compared to state-of-the-art (SOTA) ranking models like Rankformer. Additionally, it demonstrates a 1017% increase in average profit compared to all benchmark models
An introduction to generative network models and how they may be used to study animal sociality
Get PDFSocial networks constitute an important approach in the study of animal social behaviour. So far, focus has been on statistical analysis of animal social network structures. However, social networks can also be studied by generative network models - procedures that create simulated network structures. These models play a key role in wider network science, but despite occasional use, have not yet been as well integrated in the animal behaviour field. We believe that generative network models have considerable unexploited potential as a tool for understanding animal social systems. Here: 1) we provide a general introduction to generative network models, including a description of questions they are used for investigating in wider network science, explanation of key model features, and an overview of common models; 2) we consider generative network models in relation to the study of animal social behaviour, including description of questions about animal systems they can be used to investigate (demonstrated by case studies), an overview of animal behaviour studies that have used generative network modelling, the relevance of the key model features for animal behaviour studies, and consideration of how to choose a suitable generative network model for studies of animal social systems. We hope that this can help to further integrate generative network models into the study of animal sociality
In a Mirror, Dimly:Why AI Can’t Tell Our Stories and Why We Must
Get PDFToday’s generative AI tools are flooding the media ecosystem with mirrored reflections of humanity’s digitized past, reconstituted as the future. Companies are rapidly embracing these tools as ways to automate the already endangered professions of storytelling and knowledge creation. Why should we resist? After all, telling our own stories can often be painful and risky, frustrating and fruitless, or just tedious. What do we lose by surrendering the task of creating and conveying knowledge to machines that promise to remove the psychological, emotional and epistemological friction of storytelling? This article explores AI’s unwinding of the inextricable bonds among storytelling, human wisdom, knowledge and purpose, and why our future depends on their renewal
Mobilising Cultural Heritage for Locally Owned Adaptation
No full textClimate change adaptation planning and implementation has been criticised for following linear steps that can limit local suitability, scalability and sustainability. We argue that meaningful climate change adaptations incorporate a diversity of voices using cultural heritage for situated and multi-generational interventions. Here, we present examples of risk narratives and adaptive strategies developed through engagement with cultural heritage, balancing knowledge of environment with local livelihoods, histories, values and meaning
Prpf4 Sequentially Regulates the Expansion and Maturation of Erythrocyte through Distinct Mechanisms
No full textThe proliferation of early erythrocyte and the subsequent maturation are critical events during erythropoiesis, while how these two independent but interconnected processes are efficiently orchestrated during erythropoiesis is largely unknown. Prpf4 expression is enriched from Pre-Colony Forming Unit-Erythroid (PreCFU-E) to Nucleated Erythrocytes, especially in the CFU-E cells, implying that Prpf4 plays a critical role in erythropoiesis. Here, we demonstrate that prpf4 sequentially regulates erythrocyte proliferation and maturation during zebrafish definitive hematopoiesis. The data show that prpf4 mutation results in severe defects in erythropoiesis, characterized by a substantial reduction in erythroid cell numbers and impaired erythrocyte maturation. Further analysis indicates that prpf4 mutation leads to cell cycle arrest of erythrocytes at the S and G2/M phases, as well as a significant increase in erythrocyte apoptosis. Mechanistically, prpf4 mutation leads to DNA damage and the subsequent activation of the DNA damage response, triggering the ATM/CHK2-p53 signaling pathway. This process inhibits the proliferation of early erythrocyte and induces erythrocyte apoptosis. On the other hand, the data reveal that prpf4 mutation causes significant defects in skipped-exon during pre-mRNA splicing, accompanied by severe splicing defect in slc25a39 pre-mRNA. This results in a significant downregulation of slc25a39 mRNA, which partially impairs erythrocyte maturation during late erythropoiesis. In conclusion, we identify that prpf4 sequentially regulates early erythrocyte proliferation and subsequent erythrocyte maturation. This dual function of prpf4 partially explains how early erythrocyte proliferation and late maturation are efficiently coordinated during erythropoiesis
Private provision of health services in Georgia:A qualitative exploration of governance behaviours
No full textIntroductionThe private sector occupies a dominant position in Georgia’s health system, with most hospitals, primary care clinics, diagnostic facilities, pharmacies and insurance companies under for-profit ownership. Robust governance arrangements are required to align the profit-orientation of providers with health policy objectives.MethodsThis paper examines governance arrangements in Georgia’s ‘mixed’ health system. It draws on document analysis, key informant interviews, and a validation workshop. Analysis is guided by the WHO’s ‘governance behaviours’ framework, focusing on strategy, regulation, purchasing, and information-generation, as well as mechanisms for policy dialogue, actor-alignment, and trust-building. ResultsGeorgia has established a complex array of governance mechanisms for its dominant private health sector, but these remain weakly enforced. Strategic plans lack detailed implementation and budgetary integration; regulation and purchasing structures are fragmented; data systems and oversight capacity are limited; and consultation mechanisms underdeveloped - together constraining accountability, efficiency, and progress toward universal health coverage. Concluding DiscussionGeorgia’s experience highlights a persistent gap between governance intent and implementation capacity. In highly marketised systems, sustained political commitment and investment in state capacity for enforcement, data use, and stakeholder dialogue are essential to align private incentives with policy goals - and advance universal health coverage
Recovering belief structures using a language model on a naturalistic dataset of attitude change
No full textOn the Reddit forum ChangeMyView, users post beliefs and invite others to challenge them. In this study, we aimed to determine whether a GPT-4-based analytical pipeline could accurately recover belief structures from a subset of posts on predefined topics, identified through covariation statistics from a lab sample. This approach would enable us, in a second stage, to extract novel insights from naturalistic data on belief structures that have not been directly elicited in lab studies, providing a bottom-up examination at scale. Our findings suggest that the pipeline captures meaningful belief patterns, aligning moderately with human responses in structured surveys. Analyzing 3082 posts from 346 users revealed distinct ideological clusters and belief patterns that mirrored well-established political divisions. This method offers a scalable way to study belief networks, shedding light on their role in shaping societal attitudes
Navigating increasing complexity in mental health practice:ETHICA-4P,a framework and toolkit to promote reflective skills for ethical clinical decision making
No full textClinical practice for mental healthcare can be challenging. Often clinicians must make impactful decisions with little opportunity for reflection or consultation. In response to rich conversations with more than 900 local and global researchers and clinical colleagues(from more than 40 countries) about the increasing complexity of global research and clinical mental health practice, we developed the ETHICA-4P framework and toolkit for supporting clinicians in scaffolded reflection for ethical decision-making in daily practice(https://www.ethical-action.ed.ac.uk/clinical-practice). Ethical clinical interventions sit at the heart of the toolkit but practitioner-participants emphasised that ethical issues also arise during other stages of the clinical journey such as referral processes, assessment and during professional collaboration – with legacy effects long after therapy is complete. To navigate these multi layered challenges, we propose the ETHICA-4P framework and clinical toolkit centred on reflective analysis of ‘4P’s’:● Place: Social, political, cultural, and historical factors shaping practice.● People: Stakeholders involved in the ethical conflict, from clients to practitioners.● Principles: Ethical guidelines and legal frameworks.● Precedents: How past conflicts were addressed. This paper reports on the practice-driven, co-design and iterative feedback process that led to the framework and toolkit, as well as spotlighting examples of clinical applicatio
Recalibrating anti-stigma:Avoiding binary thinking and 'destigmatisation drift' in public health
No full textThis chapter argues that recalibrating stigma should involve recalibrating anti-stigma. By revisiting the ethics of stigma in public health, the authors illustrate how the moral certainty evident in the pro-/anti-stigma lobbies obstructs an acknowledgement and understanding of the complexity, inconsistency, and diversity of stigma and its affects. Using the example of anti-stigma efforts in the field of mental health, they propose a novel concept – ‘destigmatisation drift’ – to explain how approaches to anti-stigma can weaken efforts to address social drivers of suffering, illness, and even stigma itself. They continue their exploration of the tensions in anti-stigma theory and practice through the examples of ‘obesity’, anorexia, and self-harm. Each case captures how complicated the moral and practical considerations of addressing stigma are, and why an oversimplistic pro-/anti-stigma framing is an inadequate route to understanding or addressing these issues and many others
Access Repository Dashboard
Do you manage Edinburgh Research Explorer? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!