Search results

Abstract Deep learning has become an extremely important technology in numerous domains such as computer vision, natural language processing, and autonomous systems. As neural networks grow in size and complexity to meet the demands of these applications, the cost of designing and training efficient models continues to rise in computation and energy consumption. Neural Architecture Search (NAS) has emerged as a promising solution to automate the design of performant neural networks. However, conventional NAS methods often require evaluating thousands of architectures, making them extremely resource-intensive and environmentally costly.This thesis introduces a novel, energy-aware NAS pipeline that operates at the intersection of Software Engineering and Machine Learning. We present CNNGen, a domain-specific generator for convolutional architectures, combined with performance and energy predictors to drastically reduce the number of architectures that need full training. These predictors are integrated into a multi-objective genetic algorithm (NSGA-II), enabling an efficient search for architectures that balance accuracy and energy consumption.Our approach explores a variety of prediction strategies, including sequence-based models, image-based representations, and deep metric learning, to estimate model quality from partial or symbolic representations. We validate our framework across three benchmark datasets, CIFAR-10, CIFAR-100, and Fashion-MNIST, demonstrating that it can produce results comparable to state-of-the-art architectures with significantly lower computational cost. By reducing the environmental footprint of NAS while maintaining high performance, this work contributes to the growing field of Green AI and highlights the value of predictive modelling in scalable and sustainable deep learning workflows. Jury Prof. Wim Vanhoof - University of Namur, BelgiumProf. Gilles Perrouin - University of Namur, BelgiumProf. Benoit Frénay - University of Namur, BelgiumProf. Pierre-Yves Schobbens - University of Namur, BelgiumProf. Clément Quinton - University of Lille, FranceProf. Paul Temple- University of Rennes, FranceProf. Schin'ichi Satoh - National Institute of Informatics, Japan

How can teleworking and face-to-face work be reconciled? How can these professional practices be framed to reinforce the innovative and sustainable dimensions of hybrid work? These are the questions that Michel Ajzen, a specialist in organizational management, is tackling as part of his teaching assignments in the Department of Management Sciences at UNamur. His research focuses on hybrid work and organizational innovation, with a transdisciplinary approach aimed at reinventing managerial practices to meet contemporary challenges.

Floriane Goosse, a PhD student at the University of Namur, within the NaDI-CeRCLe research center, has received the prestigious "Best Research Paper Award" for her thesis paper conducted in collaboration with Wafa Hammedi, professor in the Department of Management at UNamur, and Dominik Mahr, from Maastricht University.

Our society is undergoing a digital revolution, impacting its organization, but also its practices, and even its values. Most sectors of our society have to integrate this revolution, including eHealth, eGov, eServices, collaborative economy. Solving these challenges requires a transdisciplinary approach including technology, scientific foundations, but also societal, ethical, juridical and economic viewpoints.NADI aims at federating all the UNamur researchers working on the following challenges in 7 research fields:

Building on a tradition of computer science research at the University of Namur, NaDI federates six research centers focusing on different aspects of the digital society.

Contact Co-President Bruno Dumas bruno.dumas@unamur.be Co-President Alexandre de Streel alexandre.destreel@unamur.be Organization Discover the members

NADI offers extensive expertise in artificial intelligence: bio-inspired robotics, robust, interactive, interpretable and safe machine learning, automatic program verification, declarative programming, business intelligence, knowledge representation and automatic software testing. This has already led to numerous collaborations with medical experts, industry and civil society. Along with other areas of expertise at NADI, AI experts are also exploring the educational, ethical, societal and legal implications of AI.

The collaborative economy refers to marketplaces that provide access to goods, services or skills through peer-to-peer exchanges. NADI explores the economic, technological and societal/environmental impacts of these exchanges.

Innovation has become increasingly complex. Developing appropriate solutions to our society's growing challenges requires exploring uncommon sources of solutions and combining the efforts of different stakeholders, including citizens or consumers. Co-innovation and co-creation analyze the methods, tools and governance that foster these participatory and collaborative approaches.

Japan is almost 10,000 kilometers from Belgium, a country that fascinates, not least for its rich culture full of contrasts. Researchers at UNamur maintain close ties with several Japanese institutions, particularly in the fields of computer science, mathematics and video games. Let's take a look at some of these collaborations..

This summer's third and final focus on the NaDI-CeRCLe research center, which has gained international recognition in recent weeks thanks to awards won by three young researchers in service management. Following on from Floriane Goosse and Victor Sluÿters, we invite you to discover the work of Florence Nizette, a young researcher working on Artificial Intelligence technologies.

Abstract Deep learning has become an extremely important technology in numerous domains such as computer vision, natural language processing, and autonomous systems. As neural networks grow in size and complexity to meet the demands of these applications, the cost of designing and training efficient models continues to rise in computation and energy consumption. Neural Architecture Search (NAS) has emerged as a promising solution to automate the design of performant neural networks. However, conventional NAS methods often require evaluating thousands of architectures, making them extremely resource-intensive and environmentally costly.This thesis introduces a novel, energy-aware NAS pipeline that operates at the intersection of Software Engineering and Machine Learning. We present CNNGen, a domain-specific generator for convolutional architectures, combined with performance and energy predictors to drastically reduce the number of architectures that need full training. These predictors are integrated into a multi-objective genetic algorithm (NSGA-II), enabling an efficient search for architectures that balance accuracy and energy consumption.Our approach explores a variety of prediction strategies, including sequence-based models, image-based representations, and deep metric learning, to estimate model quality from partial or symbolic representations. We validate our framework across three benchmark datasets, CIFAR-10, CIFAR-100, and Fashion-MNIST, demonstrating that it can produce results comparable to state-of-the-art architectures with significantly lower computational cost. By reducing the environmental footprint of NAS while maintaining high performance, this work contributes to the growing field of Green AI and highlights the value of predictive modelling in scalable and sustainable deep learning workflows. Jury Prof. Wim Vanhoof - University of Namur, BelgiumProf. Gilles Perrouin - University of Namur, BelgiumProf. Benoit Frénay - University of Namur, BelgiumProf. Pierre-Yves Schobbens - University of Namur, BelgiumProf. Clément Quinton - University of Lille, FranceProf. Paul Temple- University of Rennes, FranceProf. Schin'ichi Satoh - National Institute of Informatics, Japan