Tind Technologies (Norway)
Hes-so: ArODES Open Archive (University of Applied Sciences and Arts Western Switzerland / Haute école spécialisée de Suisse occidentale / FH Westschweiz)Not a member yet
15764 research outputs found
Sort by
Accruals and long-term nonfinancial assets and liabilities
No full textThis study proposes improvements to accrual models. Existing models explain how working capital maps cash flows from operations into earnings and how this mapping reflects accounting conservatism. However, except for fixed asset depreciation, accruals associated with long-term nonfinancial balance sheet accounts (e.g., intangible assets, goodwill, deferred revenues) are not modeled. We show that these unmodeled accruals have grown in importance over time and that a significant portion of them can be explained by utilizing a fundamental property of accrual accounting: most nonfinancial assets and liabilities will eventually be transferred to earnings as accruals, especially during bad times. Using a large U.S. sample for the 1988–2019 period, we document that beginning-of-year long-term nonfinancial assets and liabilities are significantly associated with total accruals and that, consistent with conditional conservatism, a greater proportion of long-term nonfinancial assets is expensed as accruals when current performance is poor. In simulations, compared to traditional models, models that include long-term nonfinancial assets and liabilities as regressors are more likely to detect seeded discretionary accruals between 2% and 20% of total assets, suggesting that these expanded models should be used to decrease the likelihood of making erroneous inferences
Future supply chains ::between new technologies and old organizational models, steering the course toward sustainability
No full textPatch-based encoder-decoder architecture for automatic transmitted light to fluorescence imaging transition ::contribution to the lightmycells challenge
No full textAutomatic prediction of fluorescently labeled organelles from label-free transmitted light input images is an important, yet difficult task. The traditional way to obtain fluorescence images is related to performing biochemical labeling which is time-consuming and costly. Therefore, an automatic algorithm to perform the task based on the label-free transmitted light microscopy could be strongly beneficial. The importance of the task motivated researchers from the FranceBioImaging to organize the LightMyCells challenge where the goal is to propose an algorithm that automatically predicts the fluorescently labeled nucleus, mitochondria, tubulin, and actin, based on the input consisting of bright field, phase contrast, or differential interference contrast microscopic images. In this work, we present the contribution of the AGHSSO team based on a carefully prepared and trained encoder-decoder deep neural network that achieves a considerable score in the challenge, being placed among the best-performing teams
Comparison between aborted/interrupted and actual suicide attempt ::An observational study on clinical and sociodemographic characteristics
No full textOverview of the 2024 ImageCLEFmedical GANs Task - investigating generative models's impact on biomedical synthetic images ::notebook for the ImageCLEF Lab at CLEF 2024
No full textThe 2024 ImageCLEFmedical GANs task Controlling the Quality of Synthetic Medical Images created via GANsis in its second edition. It comprises two sub-tasks which address the security and privacy concerns related to personal medical image data in the context of generating and using synthetic images in different real-life scenarios. The first sub-task is an extension of the task presented in the previous edition, focusing on examining the hypothesis that generative models (e.g., GANs, Diffusion Models) generate medical images containing certain “fingerprints” of the original images used for network training. The second sub-task, new this year, explores the hypothesis that generative models imprint unique fingerprints on generated images. The focus is on understanding whether different generative models or architectures leave discernible signatures within the synthetic images they produce. Ground truth data was made available to the participants. This paper presents the overview of systems and runs submitted by describing the datasets, the evaluation metrics, and discussing the methods proposed by the participating teams and their results
Quand je n'utilise pas l'IA...
No full textL'IA offre de nombreuses possibilités, mais son utilisation soulève des questions éthiques. Il est crucial d'évaluer quand l'utiliser, en tenant compte du temps, des capacités et de l'impact écologique
EsmTemp - transfer learning approach for predicting protein thermostability
No full textProtein thermostability is one of the most important features of bio-engineered proteins with significant scientific and industrial applications. Unfortunately, obtaining thermostable proteins is both expensive and complex. Recent advances in Protein Language Models (pLM) offer promising framework for sequence-to-sequence problems, especially in the realm of protein thermostability prediction. In this work, we present EsmTemp, a transfer learning model based on the ESM-2 pLM architecture. EsmTemp undergoes training on a meticulously curated dataset comprising 24,000 protein sequences with known melting temperatures. A rigorous evaluation, conducted through a 10-fold cross-validation, yields a coefficient of determination () of 0.70 and a mean absolute error of 4.3C. These outcomes highlight how pLM has the potential to advance our understanding of protein thermostability and facilitate the rational design of enzymes for various applications
BIDSAlign ::a library for automatic merging and preprocessing of multiple EEG repositories
No full textObjective. This study aims to address the challenges associated with data-driven electroencephalography (EEG) data analysis by introducing a standardised library called BIDSAlign. This library efficiently processes and merges heterogeneous EEG datasets from different sources into a common standard template. The goal of this work is to create an environment that allows to preprocess public datasets in order to provide data for the effective training of deep learning (DL) architectures. Approach. The library can handle both Brain Imaging Data Structure (BIDS) and non-BIDS datasets, allowing the user to easily preprocess multiple public datasets. It unifies the EEG recordings acquired with different settings by defining a common pipeline and a specified channel template. An array of visualisation functions is provided inside the library, together with a user-friendly graphical user interface to assist non-expert users throughout the workflow. Main results. BIDSAlign enables the effective use of public EEG datasets, providing valuable medical insights, even for non-experts in the field. Results from applying the library to datasets from OpenNeuro demonstrate its ability to extract significant medical knowledge through an end-to-end workflow, facilitating group analysis, visual comparison and statistical testing. Significance. BIDSAlign solves the lack of large EEG datasets by aligning multiple datasets to a standard template. This unlocks the potential of public EEG data for training DL models. It paves the way to promising contributions based on DL to clinical and non-clinical EEG research, offering insights that can inform neurological disease diagnosis and treatment strategies
