Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
4,798
result(s) for
"Martin, Julian"
Sort by:
Time-Aware Language Models as Temporal Knowledge Bases
Many facts come with an expiration date, from the name of the President to the basketball team Lebron James plays for. However, most language models (LMs) are trained on snapshots of data collected at a specific moment in time. This can limit their utility, especially in the closed-book setting where the pretraining corpus must contain the facts the model should memorize. We introduce a diagnostic dataset aimed at probing LMs for factual knowledge that changes over time and highlight problems with LMs at either end of the spectrum—those trained on specific slices of temporal data, as well as those trained on a wide range of temporal data. To mitigate these problems, we propose a simple technique for jointly modeling text with its timestamp. This improves memorization of seen facts from the training time period, as well as calibration on predictions about unseen facts from future time periods. We also show that models trained with temporal context can be efficiently “refreshed” as new data arrives, without the need for retraining from scratch.
Journal Article
Fabry Disease and Central Nervous System Involvement: From Big to Small, from Brain to Synapse
Fabry disease (FD) is an X-linked lysosomal storage disorder (LSD) secondary to mutations in the GLA gene that causes dysfunctional activity of lysosomal hydrolase α-galactosidase A and results in the accumulation of globotriaosylceramide (Gb3) and globotriaosylsphingosine (lyso-Gb3). The endothelial accumulation of these substrates results in injury to multiple organs, mainly the kidney, heart, brain and peripheral nervous system. The literature on FD and central nervous system involvement is scarce when focusing on alterations beyond cerebrovascular disease and is nearly absent in regard to synaptic dysfunction. In spite of that, reports have provided evidence for the CNS’ clinical implications in FD, including Parkinson’s disease, neuropsychiatric disorders and executive dysfunction. We aim to review these topics based on the current available scientific literature.
Journal Article
Psychological adjustment of Spanish adolescents and their parents during COVID-19 lockdown: A mixed method approach
Previous literature on the psychological impact of COVID-19 has shown a direct relationship between family conflicts and psychological distress among parents and their children during the domestic lockdown and social isolation; but there are also opportunities to enhance family bonding, encourage collective problem-solving and improve personal relationships. This study aimed to explore psychological adjustment processes of Spanish adolescents and their parents during the first month of lockdown by analyzing their narratives, perceived outcomes, protection and risk factors. A total of 142 people agreed to participate in this study. Of all participants, 61 were adolescents (
M
= 13.57;
SD
= 1.74; 57% women) and 81 were parents (
M
= 46.09;
SD
= 4.72; 91% mothers). All were Spanish residents and completed an online survey during the domestic lockdown in March 2020. From a qualitative design, methodology followed a mixed approach to analyze data. The results showed three different types of adaptation to lockdown and social isolation in both adolescents and their parents: 1) positive adjustment, 2) moderate adjustment, and 3) maladjustment. Most participants reported a good adjustment and only a 20% of parents and a 16% of adolescents stated that they had not been able to achieve a positive psychological adjustment. There are few significant quantitative differences between adolescents and their parents. The qualitative analysis of data showed that adolescents reported less psychological distress than their parents. The two most important protective factors were social support and keeping busy during lockdown. The most significant risk factors were loss of mobility and social isolation. The conclusions stressed that regarding psychological maladjustment, parents experienced feelings of uncertainty whereas adolescents experienced a kind of mourning process. These findings can be used to design and implement effective intervention measures for mental health and psychological well-being in such a difficult situation as domestic lockdown.
Journal Article
Targeting the Met-RIPK1 signaling axis to enforce apoptosis and necroptosis in colorectal cancer
Resistance to cell death remains a critical challenge in the therapy of colorectal cancer (CRC). Smac mimetics (SM) are cytotoxic agents specifically designed to maximize tumor cell killing mediated via endogenous tumor necrosis factor (TNF). In CRC, however, SM lacked clinical activity for reasons that remain incompletely understood. Here, we report that the clinically approved tyrosine kinase inhibitor Cabozantinib potently sensitizes to SM by targeting a Met-RIPK1 signaling axis in CRC. Aberrant Met hampers the activation of RIPK1, which renders CRC cells resistant to TNF/SM-mediated apoptosis and necroptosis. In turn, Cabozantinib potently inhibits Met, thereby stabilizing RIPK1 expression and converting TNF into a robust pro-death signal. SM/Cabozantinib-based regimens demonstrated anti-tumor activity in vivo, and were effective in a heterogeneous panel of patient-derived CRC of diverse molecular subtypes. In addition, we show that it is feasible to modulate between apoptosis and necroptosis to overcome therapy resistance and foster anti-tumor immunity. In summary, this work provides novel biological insight into the mechanisms of SM resistance and warrants the combinatory use of SM and Cabozantinib to enhance apoptotic and necroptotic cell death in CRC.
Journal Article
Determinants of Students’ Willingness to Engage in Corruption in an Academic Setting: an Empirical Study
Corruption in higher education has raised concern among governments, citizens, and the education community worldwide. However, few papers have sought to explore the students’ willingness to engage in corrupt practices at the university level. The present study aimed to examine the influence of different corrupt behaviours and perceived corruption among peers on the corrupt intention of university students. 120 undergraduate students participated in a quasi-experimental design divided in 3 treatments (control, low-corruption acceptance, high-corruption acceptance) to rate their willingness to engage in favouritism and embezzlement behaviours. Results pointed out that students were more prone to committing a non-monetary behaviour favouritism– than a monetary behaviour –embezzlement–. Furthermore, there were not significant differences between the groups of control and high-corruption acceptance; while only the group of low-corruption acceptance showed significant lower rates when compared to the control and the high-corruption acceptance’s group. Practical recommendations need to address students’ perceptions of different corrupt practices, focusing on designing ethical training programmes aimed to raise awareness on the negative consequences of non-monetary activities. Future research directions could generate empirical support to prove if students are able to recognize the underlying mechanisms of subtle corrupt practices.
Journal Article
Integrating Taxonomic, Functional and Phylogenetic Beta Diversities: Interactive Effects with the Biome and Land Use across Taxa
The spatial distribution of species, functional traits and phylogenetic relationships at both the regional and local scales provide complementary approaches to study patterns of biodiversity and help to untangle the mechanisms driving community assembly. Few studies have simultaneously considered the taxonomic (TBD), functional (FBD) and phylogenetic (PBD) facets of beta diversity. Here we analyze the associations between TBD, FBD, and PBD with the biome (representing different regional species pools) and land use, and investigate whether TBD, FBD and PBD were correlated. In the study design we considered two widely used indicator taxa (birds and ants) from two contrasting biomes (subtropical forest and grassland) and land uses (tree plantations and cropfields) in the southern Neotropics. Non-metric multidimensional scaling showed that taxonomic, functional and phylogenetic distances were associated to biome and land use; study sites grouped into four groups on the bi-dimensional space (cropfields in forest and grassland, and tree plantations in forest and grassland), and that was consistent across beta diversity facets and taxa. Mantel and PERMANOVA tests showed that TBD, FBD and PBD were positively correlated for both bird and ant assemblages; in general, partial correlations were also significant. Some of the functional traits considered here were conserved along phylogeny. Our results will contribute to the development of sound land use planning and beta diversity conservation.
Journal Article
Performance Evaluation of Open-Source Serverless Platforms for Kubernetes
Serverless computing has grown massively in popularity over the last few years, and has provided developers with a way to deploy function-sized code units without having to take care of the actual servers or deal with logging, monitoring, and scaling of their code. High-performance computing (HPC) clusters can profit from improved serverless resource sharing capabilities compared to reservation-based systems such as Slurm. However, before running self-hosted serverless platforms in HPC becomes a viable option, serverless platforms must be able to deliver a decent level of performance. Other researchers have already pointed out that there is a distinct lack of studies in the area of comparative benchmarks on serverless platforms, especially for open-source self-hosted platforms. This study takes a step towards filling this gap by systematically benchmarking two promising self-hosted Kubernetes-based serverless platforms in comparison. While the resulting benchmarks signal potential, they demonstrate that many opportunities for performance improvements in serverless computing are being left on the table.
Journal Article
The Intraday Dynamics Predictor: A TrioFlow Fusion of Convolutional Layers and Gated Recurrent Units for High-Frequency Price Movement Forecasting
This paper introduces a novel deep learning approach for intraday stock price direction prediction, motivated by the need for more accurate models to enable profitable algorithmic trading. The key problems addressed are effectively modelling complex limit order book (LOB) and order flow (OF) microstructure data and improving prediction accuracy over current state-of-the-art models. The proposed deep learning model, TrioFlow Fusion of Convolutional Layers and Gated Recurrent Units (TFF-CL-GRU), takes LOB and OF features as input and consists of convolutional layers splitting into three channels before rejoining into a Gated Recurrent Unit. Key innovations include a tailored input representation incorporating LOB and OF features across recent timestamps, a hierarchical feature-learning architecture leveraging convolutional and recurrent layers, and a model design specifically optimised for LOB and OF data. Experiments utilise a new dataset (MICEX LOB OF) with over 1.5 million LOB and OF records and the existing LOBSTER dataset. Comparative evaluation against the state-of-the-art models demonstrates significant performance improvements with the TFF-CL-GRU approach. Through simulated trading experiments, the model also demonstrates practical applicability, yielding positive returns when used for trade signals. This work contributes a new dataset, performance improvements for microstructure-based price prediction, and insights into effectively applying deep learning to financial time-series data. The results highlight the viability of data-driven deep learning techniques in algorithmic trading systems.
Journal Article
Prevalence of hepatic steatosis as assessed by controlled attenuation parameter (CAP) in subjects with metabolic risk factors in primary care. A population-based study
Primary care is the ideal setting for early identification of patients with non-alcoholic fatty liver disease (NAFLD). NAFLD is a potentially progressive disease that may lead to cirrhosis and liver cancer but is frequently underrecognized because subjects at risk are often not evaluated. Controlled attenuation parameter (CAP) is a reliable method for non-invasive quantification of liver fat. It has the advantage of simultaneous measurement of liver stiffness (LS), an estimate of liver fibrosis. There is no information on CAP in subjects with risk factors from primary care.
To investigate the prevalence of hepatic steatosis, as estimated by CAP, in subjects from the community with metabolic risk factors and correlate findings with clinical and biochemical characteristics and LS.
Population-based study of 215 subjects with metabolic risk factors without known liver disease identified randomly from a primary care center. A control group of 80 subjects matched by age and sex without metabolic risk factors was also studied. CAP and LS were assessed using Fibroscan.
Subjects with risk factors had CAP values higher than those of control group (268±64 vs 243±49dB/m,p<0.001). Prevalence of severe steatosis (CAP> 280dB/m) in subjects with risk factors was 43%. In multivariate analysis, fatty liver index (FLI) and HOMA were independent predictive factors of severe steatosis. There was a direct correlation between CAP and FLI values (r = 0.52,p<0.001). Interestingly, prevalence of increased LS was 12.6% in the risk group vs 0% in the control group (p<0.001). Increased LS occurred predominantly in subjects with high CAP values.
A high proportion of subjects with metabolic risk factors seen in primary care have severe steatosis. FLI could be used as a surrogate of CAP. Increased LS was found in a significant proportion of subjects with risk factors but not in control subjects.
Journal Article