30472 research outputs found
Sort by
SMART: Structured Missingness Analysis and Reconstruction Technique for credit scoring
The Basel Accord emphasizes the necessity of employing internal data models to manage key credit risk components, including Probability of Default (PD), Loss Given Default (LGD), and Exposure at Default (EAD). Among these, internal datasets are critical for estimating PD, a fundamental measure of borrower creditworthiness. Nevertheless, practical application often faces challenges due to incomplete datasets, which can skew analyses and undermine the accuracy of credit scoring models. Traditional approaches to addressing missing data, such as sample deletion or mean imputation, are widely used; however, they often prove insufficient for accurate prediction. Consequently, imputation methods are typically favored over deletion, as they allow for the full utilization of available data. Recent advancements have introduced more sophisticated techniques, such as Generative Adversarial Imputation Networks (GAIN), which utilize a generative adversarial network to model data distributions and impute missing values with greater precision than conventional methods. Building on these developments, this study proposes a novel imputation approach, SMART (Structured Missingness Analysis and Reconstruction Technique) for credit scoring datasets. SMART consists of two primary stages: first, it normalizes and denoises the dataset using randomized Singular Value Decomposition (rSVD), followed by the implementation of GAIN to impute missing values. Experimental results demonstrate that SMART significantly outperforms existing state-of-the-art methods, particularly in high missing data contexts (20%, 50%, and 80%), with improvements in imputation accuracy of 7.04%, 6.34%, and 13.38%, respectively. In conclusion, SMART represents a substantial advancement in handling incomplete credit scoring datasets, leading to more precise PD estimation and enhancing the robustness of credit risk management models
Cyber security management and strategic intelligence
Book synopsis: Within the organization, the cyber security manager fulfils an important and policy-oriented role. Working alongside the risk manager, the Information Technology (IT) manager, the security manager and others, the cyber security manager’s role is to ensure that intelligence and security manifest in a robust cyber security awareness programme and set of security initiatives that when implemented help strengthen the organization’s defences and those also of its supply chain partners.
Cyber Security Management and Strategic Intelligence emphasises the ways in which intelligence work can be enhanced and utilized, guiding the reader on how to deal with a range of cyber threats and strategic issues. Throughout the book, the role of the cyber security manager is central, and the work undertaken is placed in context with that undertaken by other important staff, all of whom deal with aspects of risk and need to coordinate the organization’s defences thus ensuring that a collectivist approach to cyber security management materializes. Real-world examples and cases highlight the nature and form that cyber-attacks may take and reference to the growing complexity of the situation is made clear. In addition, various initiatives are outlined that can be developed further to make the organization less vulnerable to attack. Drawing on theory and practice, the authors outline pro-active, and collectivist approaches to counteracting cyber-attacks that will enable organizations to put in place more resilient cyber security management systems, frameworks and planning processes.
Cyber Security Management and Strategic Intelligence references the policies, systems and procedures that will enable advanced undergraduate and postgraduate students, researchers, and reflective practitioners to understand the complexity associated with cyber security management and apply a strategic intelligence perspective. It will help the cyber security manager to promote cyber security awareness to a number of stakeholders and turn cyber security management initiatives into actionable policy of a proactive nature
Science for a fragile world
Imagine two worlds. In one, laws, causal relations, and mechanisms are stable. In the other, they are fragile and unreliable. Our actual world is an interlocking mixture of the two, but for many of the things we care about most, the relations that matter are fragile, and this transforms methodology. Because we cannot rely on an explanation that worked in one case still working in another, fragility requires us to re-establish empirical warrant each time. This makes the case worker central, not the theorist. And although theory is still essential, now it must be developed in a fragility-appropriate way, which means via continuous empirical refinement rather than more abstractly. A lot follows from this. Fragility tells against many uses of laboratory experiments, statistical methods, and big data, while it tells in favor of qualitative methods. It suggests a new view of scientific expertise. And the classic scientific realism debate is circumscribed because it applies only to half of science – the stable half. Fragility also tells against common practices in economics, ecology, epidemiology, and other field sciences, especially the mistaken urge to build grand, universal theories. At the same time, it calls attention to and lionizes much other, ingenious work in these fields that is effective despite difficult circumstances. A philosophy of science attuned to fragility gives us a new handle on science
Confirmation bias and the plausible distribution of evidence
This paper discusses different, inter-related notions of confirmation bias as clas-
sified by Nickerson (1998), distinguishing broadly between confirmation bias in
evidence search, evidence selection, evidence evaluation, and evidence communi-
cation. It then sets out an overlooked rational basis for seeming bias in evidence
search that stems from reasonable expectations about the real-world distribution
of evidence in favour and against a claim. The implications of this for commu-
nication and argumentation are discussed, as are the implications of taking into
account plausible evidence distributions for research on bias more generally
Assessing national technological advantage in circular economy innovations for the textile and leather sectors: a patent-based analysis
This paper aims to measure and monitor countries’ technological advantage in the circular economy related technologies in the textile and leather industry. Analysing USPTO patent data, first, we identified the patent classes which are more likely to include the technologies for circular economy in the textile and leather sector. Second, we identified how the capabilities relevant in these areas are distributed across nations. Our results show that Italy still lags behind other European and non-European countries in terms of technological capabilities for circular economy for the areas specific to the textile and leather sector. The results allow us to identify some of the national priorities to foster the competitiveness of one of the pillars of the Made in Italy industries
How can we achieve sustainable funding for open access books?
Is the biggest blocker to open access (OA) for books actually the economics of it all? Book processing charges (BPCs) do not scale but they remain a significant method of paying to produce OA monographs for many researchers and libraries. However, in the last few years, we have seen several new initiatives emerge that seek to solve the problem posed by funding via BPCs alone. There is a proliferation of collective funding models for OA books, including Opening the Future, Open Book Collective, MIT Press’s D2O, JSTOR’s Path to Open and others. They all work differently, but they all offer alternatives to BPCs. In this article we explore the theme of sustainable funding for OA monographs, presenting a range of new models, and suggest that their normalization is well overdue. We also present the work of the library at Lancaster University on their new strategy supporting open access. While this article takes a somewhat UK-centric path, what is happening in the UK may be replicated in other countries and contexts. With demand increasing for monographs to be open this is a timely topic. The authors welcome discussion from publishers, libraries and other stakeholders
What to do when your Tor relay isn't listening
I had a problem with my Tor relay last night. For some reason, the application (daemon) started but then after about 5 seconds it stopped listening. The Tor Metrics site was displaying a red button saying my node was unreachable and when I did “sudo ss -ltp” I got nothing. The application literally wasn’t “listening” (in network socket terms).
The answer was actually embarrassingly simple. If your Tor daemon stops listening, even if you can’t find a log entry saying why, it could be because you have burned through your entire traffic allocation for the accounting period. Increase the limit and try restarting it to avoid a red face
How to securely create an encrypted digital 'in case of death' document
Very few people like thinking about the fact they will die. But it can prove a substantial administrative headache to loved ones if they don’t know about all of your finances, your passwords, your emails etc. At the same time, you shouldn’t be writing down passwords in any document that could be stolen or seen; it’s bad cybersecurity practice. So what can you do? This post details the technicalities
Rethinking the death penalty through a public health lens: examples from Malaysia
This paper examines the death penalty through a public health framework, analysing its systemic impacts on individual and
community well-being. Drawing on recent legislative changes in Malaysia, including the 2023 abolition of mandatory capital
punishment, the study identi es three critical pathways through which the death penalty threatens public health: prolonged
con nement of death row inmates, miscarriages of justice, and punitive drug policies. The research highlights how capital pun-
ishment creates trauma that extends beyond death row prisoners and affects families, prison staff, legal professionals, and
communities. It emphasises the death penalty’s disproportionate impact on marginalised groups, including women, sexual
minorities, and individuals with disabilities, who face compounded discrimination in the judicial system. The paper concludes
that abolishing the death penalty is not merely a legal imperative but a public health necessity, recommending a shift toward
evidence-based, health-centred approaches that prioritise rehabilitation, harm reduction, and restorative justice