9095 research outputs found
Sort by
Detecting Financial Statement Fraud: An Alternative Evaluation of Automated Tools Using Portfolio Performance
This article investigates the effect of using financial statement fraud detection models in constructing investment portfolios. Three financial statement fraud detection models are recreated and used to inform portfolio construction. Portfolio performance is compared between two strategies investing in companies on the S&P 500 predicted to have the highest (lowest) likelihood of financial statement fraud according to three models. Investment performance under the two strategies and across the three models are assessed using Fama-French regressions over a trading period from 2003 to 2021 and during market shocks. The portfolio of companies with the highest likelihood of fraud underperforms, characterized by inadequate returns relative to risk exposures. In the case of low-likelihood firms, results are consistent with risk-reward expectations. Financial results were consistent across all three fraud models, indicating that each model effectively discriminates between companies predicted to exhibit financial statement fraud. This research investigates the effect of financial statement fraud risk on investment performance and provides an alternative evaluation of financial statement fraud detection models, complementing the traditional accounting analysis of such models
Data-Driven Erbium-doped Fiber Amplifier Gain Modeling Using Gaussian Process Regression
We propose a data-driven erbium-doped fiber amplifier (EDFA) gain model utilizing Gaussian process regression (GPR). An additive Laplacian and radial-basis function kernel is proposed for the GPR and was found to outperform deep neural network (DNN) methods while additionally providing prediction uncertainty. Performance is measured using mean absolute error (MAE) averaged across five different EDFAs with three manufacturers. The GPR achieves an MAE of 0.1 dB using 30 training samples in contrast to the DNN that achieves an MAE of 0.25 dB using 3000 training samples. Additionally, we demonstrate that active learning can be used to improverobustness and repeatability of convergence
Physical activity in adult users of inpatient mental health services: A scoping review
People with severe mental illness engage in considerably less physical activity than those without. They also experience premature mortality of around 10–25 years. A large proportion of these premature deaths are attributed to modifiable behaviours, including physical activity. The inpatient environment provides an opportunity to support people to become more physically active; however, there is limited evidence on which interventions are most successful and what contextual factors affect their delivery. A scoping review was conducted to help understand the extent and type of evidence in this area and identify research gaps. We included studies of physical activity correlates and interventions in adult inpatient mental health services published in peer-reviewed journals. Reviews, meta-analyses, and papers focusing on eating disorder populations were excluded. We searched the MEDLINE, CINAHL, PsycINFO, ASSIA and Web of Science databases for relevant studies published in English. We extracted data on study design, participant characteristics, intervention and control conditions, key findings, and research recommendations. We used a descriptive analytical approach and results are presented in tables and figures. Of 27,286 unique records screened, 210 reports from 182 studies were included. Sixty-one studies reported on correlates of physical activity, and 139 studies reported on physical activity interventions. Most intervention studies used a single-group, pre-post design (40%) and included fewer than 100 participants (86%). Ninety percent of interventions delivered physical activity directly to participants, and 50% included group-based sessions. The duration, type, frequency and intensity of sessions varied. Mental health was the most commonly reported outcome (64%), whereas physical activity was rarely an outcome (13%). Overall, there is a modest but growing body of research on physical activity in adult users of inpatient mental health services. More high-quality trials are needed to advance the field, and future research should target neglected intervention types, outcomes, populations and settings
Breeding habitat loss linked to declines in Rufous Hummingbirds
Habitat loss is the primary driver of biodiversity decline worldwide, but it remains unknown how land-cover change and, in general, habitat loss impact many migratory species, such as the Rufous Hummingbird (Selasphorus rufus). Here, we gathered 5115 occurrence records for the Rufous Hummingbird from professional and citizen-science data sets and parameterized species distribution models with four bioclimatic variables and two Landsat satellite spectral reflectance bands. We calculated the population change and change in the potential distribution of the Rufous Hummingbird across its breeding range in the Pacific Northwest of North America over the last 36 yr (1985–2021). Back-casting habitat suitability predictions over time, we provide the first quantifications of breeding habitat change for the Rufous Hummingbird, which has exhibited precipitous declines over the past two decades. Furthermore, we evaluated links between modeled habitat suitability, population abundance, and trends with a route-level analysis of Breeding Bird Survey data. We found notable habitat loss occurring in Bird Conservation Regions along the Pacific coast where the species is most abundant (54% and 34% decreases in suitable habitat area), with habitat loss in coastal regions linked to population decline. In contrast, we detected habitat gains in regions along the interior, northeastern edges of the breeding range (160% and 85% increases in suitable habitat area). However, increasing suitability does not guarantee species colonization of new habitat. Our results indicate the need to further investigate drivers of habitat loss, such as intensive forestry and suppression of early seral habitat, along the Pacific coast. Our modeling approach can be applied to efficiently detect and quantify habitat loss over time for a variety of taxa
Social psychiatry in Oxford and its ecological niche, 1959–1988
This paper reflects on a special edition of the journal History of Psychiatry and a related symposium held at Somerville College, Oxford, exploring the innovations in mental healthcare in Oxfordshire led by Dr Bertram Mandelbrote between 1959 and 1988. I draw on clinical culture, biography, mental health policy and my lived experience to understand Mandelbrote's life and work, and his legacy and lessons for contemporary psychiatrists. I explore the ecological niche that Mandelbrote created and conclude with the probable importance of his relationship with Professor Michael Gelder, who led the University of Oxford Department of Psychiatry at the time
Representation of Sex from the Face and Body: Evidence from a Visual Adaptation Task
Observers readily make judgments of sex from the appearance of others' faces and bodies. Researchers have used visual adaptation techniques with such social stimuli to demonstrate high level after-effects such that ambiguous bodies or faces appear more male or female following adaptation to a female or male adaptor. Such findings have typically been interpreted in terms of a symmetrical mental 'space'. However, previous evidence from visual search tasks suggests that ‘female’, in both body shape and in face appearance, is coded as an extension of a ‘male’ default. That is, there is an asymmetry in the representation of ‘male’ and ‘female’ in the perceptual system. This hypothetical polarised representation of sex predicts asymmetric effects of adaptation in adaptation paradigms. However, to our knowledge, there are no attempts in the literature to directly compare the strength of the perceptual aftereffect of adaptation for ‘male’ versus ‘female’ faces and bodies. To this end, we utilised a visual adaptation paradigm, where participants were exposed to computer-generated strongly female, strongly male, or perceptually ambiguous faces and body silhouettes (adaptors) and were then asked to categorise subsequent slightly male, slightly female, or perceptually ambiguous faces or body silhouettes. The results showed a strong aftereffect of adaptation, such that after prolonged exposure to strongly ‘male’ stimuli, participants categorized the target stimuli as more ‘female’, and vice versa. However, to date, we have not found strong evidence for asymmetric representation of ‘male’ compared to ‘female’, contrary to our original hypothesis. Evidence about the organisation of male/female face and body representations does not converge on whether they are symmetrical or asymmetrical
Language and Technology in Wales: Volume II
This ebook is based on some of the main papers presented during the Welsh Language Technologies Academic Symposium 2022 and 2023. It contains chapters dealing with the latest technology in the field of Artificial Intelligence (AI), natural language processing (NLP) and speech technology, techniques that continue to offer much to less-resourced languages. In addition to reporting on the latest technology available for the Welsh language, included are discussions on language technology tools for languages that are in a similar situation, namely Breton, Cornish and Irish
Threshold Gain Reduction in Tandem Semiconductor Nano-Lasers
It is shown that a significant reduction in the threshold gain of electrically pumped semiconductor nano-lasers may be achieved in bridge-connected tandem semiconductor nano-lasers. Optimization of the design is achieved by exploring the impact of bridge length and width on the threshold gain. In addition, a detailed examination is also made of the emission patterns of the structure. It is found that a trade-off emerges between threshold gain and beam quality where multi-lobed far field emission may be associated with the lowest threshold gains
Assessing Li accommodation at amorphous ZrO2 grain boundaries
Nuclear Pressurised Water Reactors (PWRs) use zirconium alloys as a fuel cladding, preventing the cooling water, at elevated pH using lithium hydroxide, from interacting with the fuel. Boron, as boric acid, is added to the coolant as a reactivity shim. Future reactor designs are considering removing soluble boron reactivity control to aid plant simplification. The presence of lithium in the absence of boron in the coolant has, however, been found to accelerate the corrosion of zirconium-based alloys under certain conditions and the mechanisms by which this occurs is under investigation. The ingress of lithium into the bulk oxide layer of zirconium alloy has been addressed in a previous study and was found to be unlikely. Here, atomistic simulations were used to produce Brouwer diagrams from which the solubility of lithium in amorphous structures representing complex grain boundaries have been predicted. The solubility of lithium in these amorphous structures is predicted to be high and will produce an elevated concentration of oxygen defects within the amorphous structure. This could offer a mode for transport of oxygen to the metal oxide interface and, potentially, offer a mechanism or part of a mechanism for observed lithium-accelerated corrosion of Zr-based alloys