Résultats de recherche

Abstract In recent decades, the volume of data generated worldwide has grown exponentially, significantly accelerating advancements in machine learning. This explosion of data has led to an increased need for effective data exploration techniques, giving rise to a specialized field known as dimensionality reduction. Dimensionality reduction methods are used to transform high-dimensional data into a low-dimensional space (typically 2D or 3D), so that it can be easily visualized and understood by humans. Algorithms such as Principal Component Analysis (PCA), Multidimensional Scaling (MDS), and t-distributed Stochastic Neighbor Embedding (t-SNE) have become essential tools for visualizing complex datasets. These techniques play a critical role in exploratory data analysis and in interpreting complex models like Convolutional Neural Networks (CNNs). Despite their widespread adoption, dimensionality reduction techniques, particularly non-linear ones, often lack interpretability. This opacity makes it difficult for users to understand the meaning of the visualizations or the rationale behind specific low-dimensional representations. In contrast, the field of supervised machine learning has seen significant progress in explainable AI (XAI), which aims to clarify model decisions, especially in high-stakes scenarios. While many post-hoc explanation tools have been developed to interpret the outputs of supervised models, there is still a notable gap in methods for explaining the results of dimensionality reduction techniques. This research investigates how post-hoc explanation techniques can be integrated into dimensionality reduction algorithms to improve user understanding of the resulting visualizations. Specifically, it explores how interpretability methods originally developed for supervised learning can be adapted to explain the behavior of non-linear dimensionality reduction algorithms. Additionally, this work examines whether the integration of post-hoc explanations can enhance the overall effectiveness of data exploration. As these tools are intended for end-users, we also design and evaluate an interactive system that incorporates explanatory mechanisms. We argue that combining interpretability with interactivity significantly improves users' understanding of embeddings produced by non-linear dimensionality reduction techniques. In this research, we propose enhancements to an existing post-hoc explanation method that adapts LIME for t-SNE. We introduce a globally-local framework for fast and scalable explanations of t-SNE embeddings. Furthermore, we present a completely new approach that adapts saliency map-based explanations to locally interpret non-linear dimensionality reduction results. Lastly, we introduce our interactive tool, Insight-SNE, which integrates our gradient-based explanation method and enables users to explore low-dimensional embeddings through direct interaction with the explanations. Jury Prof. Wim Vanhoof - University of Namur, BelgiumProf. Benoit Frénay - University of Namur, BelgiumProf. Bruno Dumas - University of Namur, BelgiumProf. John Lee - University of Louvain, BelgiumProf. Luis Galarraga - University of Rennes, France

Une équipe internationale de chercheurs vient de publier dans la prestigieuse revue Light : Science & Applications (LSA) du groupe Nature.  Les équipes des professeurs Michaël Lobet et Alexandre Mayer (Université de Namur) ont collaboré avec l’équipe du professeur Shanhui Fan, l’un des plus grands spécialistes en la matière, de la prestigieuse Université de Stanford, en Californie (USA).  Le résultat : un article intitulé « Twist-Induced Beam Steering and Blazing Effects in Photonic Crystal Devices” ou l’étude de la déviation d’un faisceau par torsion dans les dispositifs à cristaux photoniques.  Allez, on re-twiste à l’UNamur ! 

Plusieurs chercheurs du Namur Institute of Structured Matter (NISM) ont récemment obtenu des financements du F.R.S – FNRS à la suite des appels dont les résultats ont été publiés en décembre 2024.  L’Institut NISM fédère les activités de recherche des départements de chimie et de physique de l’Université de Namur. 

Comment concilier télétravail et travail en présentiel ? Comment encadrer ces pratiques professionnelles pour renforcer les dimensions innovantes et durables du travail hybride ? C’est à toutes ces questions que Michel Ajzen, spécialiste en management des organisations, s’intéresse dans le cadre de ses missions d’enseignement au sein du département des sciences de gestion de l’UNamur. Ses recherches se concentrent sur le travail hybride et l'innovation organisationnelle, avec une approche transdisciplinaire visant à réinventer les pratiques managériales pour relever les défis contemporains. 

Floriane Goosse, doctorante à l’Université de Namur, au sein du centre de recherche NaDI-CeRCLe, a reçu le prestigieux prix « Best Research Paper Award » pour son papier de thèse mené en collaboration avec Wafa Hammedi, professeure au Département de gestion de l’UNamur, et Dominik Mahr, de l’Université de Maastricht. 

Programme 8:50 | Welcome, registration, and poster setup9:20 | Welcome speechFirst morning session 9:30 | Vladimir N. Uversky - University of South Florida, USA - "Dancing protein clouds: strange biology and chaotic physics of intrinsically disordered proteins" 10:15 | Marie Skepö - Lunds Universitet, Sweden - "Structural and conformation properties of IDPs: computer simulations in combination with experiments"       11:00 | Coffee breakSecond morning session11:30 | Peter Tompa - Vrije Universiteit Brussel, Belgium - "Fuzzy interactions of IDPs driving biomolecular condensation"12:15 | Sonia Longhi - Aix-Marseille Université, France - "Intrinsic disorder, phase transitions, and fibril formation by the Henipavirus V and W proteins"   13:00 | Lunch and poster sessionAfternoon session14:30 | Sigrid Milles - Leibniz-Forschungsinstitut für Molekulare Pharmakologie, Germany - "Intrinsically disordered proteins in endocytosis: an NMR and single molecule fluorescence perspective"     15:15 | Jean-François Collet - Université Catholique de Louvain, Belgium - "How disorder controls the transport of lipoproteins in the cell envelope of Gram-negative bacteria"    16:00 | Closing speech16:10 | Networking Beer Time at "Le Chapitre"VenueUniversity of Namur, auditorium Pedro Arrupe (PA02), rue de Bruxelles, 65-67 - 5000 Namur (#21 on the campus map) Download the programme (PDF) Download the campus map Registration guidelines Registration feeStudents (PhD students included): 25 €Seniors: 40 €Payment - Bank transferPayable before 6 December on the account:Name: Université de Namur – ASBLIBAN: BE10 2500 0740 2704BIC: GEBABEBB Please mention your name/CPO4136330 /e-mail in the payment communication. Abstract guidelines Send us your abstract before 6 December by email: pdid.meeting@unamur.be Format: Word document, maximum 1 page A4, Times New Roman  Registration All deadlines (registration, payment, abstracts) : 6 December 2024