Search

Your search keyword '"De Paepe, P."' showing total 999 results
Start Over Author "De Paepe, P."
999 results on '"De Paepe, P."'

1. Joint Reconstruction of the Activity and the Attenuation in PET by Diffusion Posterior Sampling: a Feasibility Study

Author

Phung-Ngoc, Clémentine, Bousse, Alexandre, De Paepe, Antoine, Dang, Hong-Phuong, Saut, Olivier, and Visvikis, Dimitris

Subjects
Physics - Medical Physics
Abstract

This study introduces a novel framework for joint reconstruction of the activity and attenuation (JRAA) in positron emission tomography (PET) using diffusion posterior sampling (DPS), which leverages diffusion models (DMs) to address activity-attenuation dependencies and mitigate crosstalk issues in non-time-of-flight (TOF) settings; experimental evaluations with 2-dimensional (2-D) XCAT phantom data show that DPS outperforms traditional maximum likelihood activity and attenuation (MLAA) methods, providing consistent, high-quality reconstructions even without TOF information., Comment: 5 pages, 4 figures

Published
2024

3. DIRESA, a distance-preserving nonlinear dimension reduction technique based on regularized autoencoders

Author

De Paepe, Geert and De Cruz, Lesley

Subjects
Computer Science - Machine Learning, Nonlinear Sciences - Chaotic Dynamics, Physics - Atmospheric and Oceanic Physics
Abstract

In meteorology, finding similar weather patterns or analogs in historical datasets can be useful for data assimilation, forecasting, and postprocessing. In climate science, analogs in historical and climate projection data are used for attribution and impact studies. However, most of the time, those large weather and climate datasets are nearline. They must be downloaded, which takes a lot of bandwidth and disk space, before the computationally expensive search can be executed. We propose a dimension reduction technique based on autoencoder (AE) neural networks to compress those datasets and perform the search in an interpretable, compressed latent space. A distance-regularized Siamese twin autoencoder (DIRESA) architecture is designed to preserve distance in latent space while capturing the nonlinearities in the datasets. Using conceptual climate models of different complexities, we show that the latent components thus obtained provide physical insight into the dominant modes of variability in the system. Compressing datasets with DIRESA reduces the online storage and keeps the latent components uncorrelated, while the distance (ordering) preservation and reconstruction fidelity robustly outperform Principal Component Analysis (PCA) and other dimension reduction techniques such as UMAP or variational autoencoders., Comment: 16 pages, 10 figures, 4 tables; 7 pages of Supporting Information

Published
2024

16. C-terminal frameshift variant of TDP-43 with pronounced aggregation-propensity causes rimmed vacuole myopathy but not ALS/FTD

Author

Ervilha Pereira, Pedro, Schuermans, Nika, Meylemans, Antoon, LeBlanc, Pontus, Versluys, Lauren, Copley, Katie E., Rubien, Jack D., Altheimer, Christopher, Peetermans, Myra, Debackere, Elke, Vanakker, Olivier, Janssens, Sandra, Baets, Jonathan, Verhoeven, Kristof, Lammens, Martin, Symoens, Sofie, De Paepe, Boel, Barmada, Sami J., Shorter, James, De Bleecker, Jan L., Bogaert, Elke, and Dermaut, Bart

Published
2023
Full Text
View/download PDF

17. Deep Learning for Digital Asset Limit Order Books

Author

Jha, Rakshit, De Paepe, Mattijs, Holt, Samuel, West, James, and Ng, Shaun

Subjects
Quantitative Finance - Statistical Finance
Abstract

This paper shows that temporal CNNs accurately predict bitcoin spot price movements from limit order book data. On a 2 second prediction time horizon we achieve 71\% walk-forward accuracy on the popular cryptocurrency exchange coinbase. Our model can be trained in less than a day on commodity GPUs which could be installed into colocation centers allowing for model sync with existing faster orderbook prediction models. We provide source code and data at https://github.com/Globe-Research/deep-orderbook., Comment: 9 pages, 7 figures

Published
2020

28. Learner Characteristics, Learner Achievement and Time Investment in Online Courses for Dutch L2 in Adult Education

Author

De Paepe, Liesbeth, Zhu, Chang, and Depryck, Koen

Abstract

For the growing group of adult migrants, flexible solutions for second language (L2) acquisition are increasingly important, while concerns about the efficacy of online language learning abound. This study on the L2 situation in Flanders has 4 key aims: analyzing adult learner profiles in fully online Dutch beginners' courses; comparing learner achievements in fully online and face-to-face (F2F) courses; reporting on differences in learner achievement considering learner profile variables and comparing time investment in F2F and online courses. A quantitative design was adopted, including a survey among the online students; assessment scores and data on learner profiles of online and F2F learners. The survey dealt with biographical variables and variables related to online learning. Results show that i) the online Dutch L2 learner is typically highly educated and employed; ii) online learners spend less time on the course than F2F learners, but perform significantly better in reading, listening and speaking skills; iii) writing skills are better achieved in F2F learning and iv) prior educational level and the language spoken by the host community have no significant impact on learner achievement. The findings indicate that online L2 learning can be as effective as F2F learning, even for learners with a lower educational level, or lacking contact with native speakers.

Published
2018

29. Student Performance in Online and Face-to-Face Second Language Courses: Dutch L2 in Adult Education

Author

de Paepe, Liesbeth

Abstract

Due to immigration, globalization and a changing labor market, there is a recognized need for flexible learning. In Europe, second language (L2) acquisition specifically has become of vital importance. Research on less commonly taught languages (LCTL), however, is scarce and focuses mainly on foreign language acquisition, compulsory education and on face-to-face (F2F) and blended learning. However, the effectiveness of learning a LCTL -- such as Dutch -- as a second language fully online in adult education has yet to be uncovered. The principal objective of this paper was to investigate whether learning Dutch L2 fully online can be as effective as learning it fully F2F. A quantitative methodological approach was adopted. Independent samples t-tests were carried out to compare online and F2F adult learners' performance in a Waystage-level Dutch L2 course in Flanders, Belgium. The key variables 'previous knowledge', 'course level', 'content', 'teacher', 'assessment' and 'learning outcomes' are controlled. Main findings indicated that online Dutch L2 learning in adult education can be at least as effective as F2F learning. This study should, therefore, be of value to language course providers wishing to implement flexible L2 learning, and to LCTL and computer-assisted language learning (CALL) researchers.

Published
2018

36. The long non-coding RNA SAMMSON is essential for uveal melanoma cell survival

Author

Dewaele, Shanna, Delhaye, Louis, De Paepe, Boel, de Bony, Eric James, De Wilde, Jilke, Vanderheyden, Katrien, Anckaert, Jasper, Yigit, Nurten, Nuytens, Justine, Vanden Eynde, Eveline, Smet, Joél, Verschoore, Maxime, Nemati, Fariba, Decaudin, Didier, Rodrigues, Manuel, Zhao, Peihua, Jochemsen, Aart, Leucci, Eleonora, Vandesompele, Jo, Van Dorpe, Jo, Marine, Jean-Christophe, Van Coster, Rudy, Eyckerman, Sven, and Mestdagh, Pieter

Published
2022
Full Text
View/download PDF

40. E-Content Development for Languages: Success Factors and Pitfalls

Author

De Paepe, Liesbeth

Abstract

This paper discusses the success factors and pitfalls in development of e-content for languages. The factors discussed draw on several years of experience in developing and implementing 95% distance courses for Dutch as a second language in the adult education sector in Flanders and on PhD research at VUB. The CEFR [Common European Framework of Reference for Languages] provides language centres with a profound pedagogical framework necessary when developing language courses. The 10 success factors and 5 pitfalls in the development of e-content for languages discussed in this text comprise practical guidelines, which can affect the process of creating qualitative e-content for languages positively.

Published
2014

48. Development and Implementation of Online Dutch L2 Courses in Adult Education: Educators' and Providers' Perceptions of Constraints and Critical Success Factors

Author

De Paepe, Liesbeth, Zhu, Chang, and Depryck, Koen

Abstract

This study examines educators' and providers' perceptions of constraints and critical success factors of the development and implementation of online L2 learning. The focus is on Dutch L2 learning in adult education in Flanders. Semi-structured interviews with seven educators (shaping pedagogies/curricula) and six providers (organizing/offering language programmes) were conducted. Consequently, a focus group consisting of 12 teachers of Dutch (L2) and teacher trainers was run. The findings show that the main perceived constraints of the design and implementation of an online curriculum are negative beliefs about the effectiveness of online language learning, costs, a lack of technical and pedagogical support and insufficient teacher ICT-skills. The perceived major critical success factors are course design, technological facilities, (technical, financial and pedagogical) support, learner skills and attitudes, and competencies of instructional designers. Although many literature references were found concerning the importance of assessment, teacher skills and the role of the faculties, these were hardly mentioned during the interviews. The findings of this study call for a better financial framework, more collaboration among different stakeholders in the field, a focus on online teacher skills, adequate pedagogical and technical support and for a framework for quality assurance in online language learning.

Published
2019
Full Text
View/download PDF

49. Drop-Out, Retention, Satisfaction and Attainment of Online Learners of Dutch in Adult Education

Author

De Paepe, Liesbeth, Zhu, Chang, and DePryck, Koen

Abstract

Understanding reasons for drop-out, factors contributing to student satisfaction and strategies for enhancing student retention is fundamental to the success of online (language) courses. This study examines drop-out rates, student satisfaction, retention rates and attainment in fully online elementary-level courses for Dutch as a second language (L2) in Flanders. Results show that drop-out rates are highly "front-loaded" and that success in the A1 level course is a good predictor for success in the subsequent level. Retention strongly increases after the first module. Student-tutor interaction is a key element in student satisfaction that deserves attention from instructional designers, policy makers and educators.

Published
2018

Catalog

Books, media, physical & digital resources
See catalog results