Results 1-12 of 12.
Cuvelier+Thibaut+

Bookmark and Share    
See detailCharacterising Industrial Sites' Flexibility with Reservoir Models
Cuvelier, Thibaut ULiege

Poster (2017, August)

Electro-intensive industrial sites are very dependent on electricity prices to remain competitive. Nevertheless, they can often tune their processes in order to decrease their electricity consumption ... [more ▼]

Electro-intensive industrial sites are very dependent on electricity prices to remain competitive. Nevertheless, they can often tune their processes in order to decrease their electricity consumption during the most critical periods, for example by using decision support systems based on mathematical modelling of their processes. Our goal is to estimate the flexibility potential of a complete site, not to tune each process very precisely. To this end, we propose a generic paradigm to help conceiving such models: reservoirs are the basic building block, which allows for great expressiveness while being close to the physics. More specifically, we do not need very precise models for our purposes, but ones that can be efficiently included in optimisation models. Our first results show that the obtained reservoir models can give sufficiently good approximations for metallurgical and other processes. [less ▲]

Detailed reference viewed: 27 (5 ULiège)
Full Text
Peer Reviewed
See detailNETPerfTrace – Predicting Internet Path Dynamics and Performance with Machine Learning
Wassermann, Sarah ULiege; Casas, Pedro; Cuvelier, Thibaut ULiege et al

in Proceedings of Big-DAMA ’17 (2017, August)

We study the problem of predicting Internet path changes and path performance using traceroute measurements and machine learning models. Path changes are frequently linked to path inflation and ... [more ▼]

We study the problem of predicting Internet path changes and path performance using traceroute measurements and machine learning models. Path changes are frequently linked to path inflation and performance degradation, therefore the relevance of the problem. We introduce NETPerfTrace, an Internet Path Tracking system to forecast path changes and path latency variations. By relying on decision trees and using empirical distribution-based input features, we show that NETPerfTrace can predict (i) the remaining life time of a path before it actually changes and (ii) the number of path changes in a certain time period with relatively high accuracy. Through extensive evaluation, we demonstrate that NETPerfTrace highly outperforms DTRACK, a previous system with the same prediction targets. NETPerfTrace also offers path performance forecasting capabilities. In particular, our tool can predict path latency metrics, providing a system which can not only predict path changes, but also forecast their impact in terms of performance variations. We release NETPerfTrace as open software to the networking community, as well as all evaluation datasets. [less ▲]

Detailed reference viewed: 97 (16 ULiège)
Full Text
Peer Reviewed
See detailOptimising workforce and energy costs by exploiting production flexibility
Cuvelier, Thibaut ULiege; Louveaux, Quentin ULiege

Conference (2017, July)

In a world where the electricity prices become more and more volatile, notably due to renewable energies, the industry is suffering from cost variations never seen before, especially when electro ... [more ▼]

In a world where the electricity prices become more and more volatile, notably due to renewable energies, the industry is suffering from cost variations never seen before, especially when electro-intensive. Nevertheless, the plants can significantly reduce this impact: some electro-intensive factories could shift their production to time periods where the electricity is cheaper, resulting in large savings. At the same time, the grid operator can remunerate this consumption adaptation as a flexibility service. Our research goal is to optimise the operations of a factory around this flexibility. We compute a production plan that adapts to price forecasts, but also flexibility levers that adjust this plan to react to unexpected price changes. We propose the unifying concept of reservoir to provide sufficiently good models for the plant's processes. Nevertheless, this methodology implies to have frequent production plan changes, which directly impacts the workers, as they may be asked to follow barely predictable schedules. This has a significant detrimental effect on their quality of life. As a consequence, the human aspect of flexibility must also be considered: we seek for production plans that consider both workforce and energy costs, and we then assign workers to work shifts while ensuring their well-being. This HR orientation is the most innovative contribution of this research project. [less ▲]

Detailed reference viewed: 88 (18 ULiège)
Full Text
See detailA Journey Through Julia: A dynamic and fast language
Cuvelier, Thibaut ULiege

Speech/Talk (2017)

Usually, dynamic programming languages (like Python, R, or MATLAB) are quite slow when executed, which causes performance problems in many applications. Julia is a blossoming language, both dynamic and ... [more ▼]

Usually, dynamic programming languages (like Python, R, or MATLAB) are quite slow when executed, which causes performance problems in many applications. Julia is a blossoming language, both dynamic and fast, fully open source, with a syntax that is very similar to that of MATLAB. This talk presents briefly Julia and its ecosystem. [less ▲]

Detailed reference viewed: 14 (2 ULiège)
Full Text
See detailPredicting Internet Path Dynamics and Performance with Machine Learning
Wassermann, Sarah ULiege; Casas, Pedro; Cuvelier, Thibaut ULiege et al

Report (2017)

In this paper, we study the problem of predicting Internet path changes and path performance using traceroute measurements and machine learning models. Path changes are frequently linked to path inflation ... [more ▼]

In this paper, we study the problem of predicting Internet path changes and path performance using traceroute measurements and machine learning models. Path changes are frequently linked to path inflation and performance degradation; therefore, predicting their occurrence is highly relevant for performance monitoring and dynamic traffic engineering. We introduce NETPerfTrace, an Internet Path Tracking system capable of forecasting path changes and path latency variations. By relying on decision trees and using empirical distribution based input features, we show that NETPerfTrace can predict (i) the remaining life time of a path before it actually changes and (ii) the number of path changes in a certain time-slot with high accuracy. Through extensive evaluation, we demonstrate that NETPerfTrace highly outperforms DTRACK, a previous system with the same prediction targets. NETPerfTrace also offers path performance forecasting capabilities. In particular, it can predict path latency metrics, providing a system which could not only predict path changes but also forecast their impact in terms of performance variations. As an additional contribution, we release NETPerfTrace as open software to the networking community. [less ▲]

Detailed reference viewed: 90 (9 ULiège)
See detailCréer des applications graphiques en Python avec PyQt5
Denis, Pierre; Cuvelier, Thibaut ULiege

Book published by D-BookeR - 1st (2017)

Ce livre s'adresse à toute personne souhaitant développer des interfaces graphiques en Python, de la simple boîte de dialogue à la création d'applications graphiques élaborées, avec menus, barre d'outils ... [more ▼]

Ce livre s'adresse à toute personne souhaitant développer des interfaces graphiques en Python, de la simple boîte de dialogue à la création d'applications graphiques élaborées, avec menus, barre d'outils, redimensionnement à la taille de l'écran, accès à une base de données, etc. Il part du principe que vous connaissez Python et avez des notions en programmation orientée objet, mais que vous débutez avec Qt et la conception d'interfaces graphiques. Qt est une bibliothèque multiplateforme extrêmement complète écrite en C++. PyQt sert de couche de liaison entre ces deux mondes et apporte Qt à l'environnement Python. Ce livre s'appuie sur la version 5.6 de Qt, destinée à être maintenue sur le long terme, tout en signalant le cas échéant d'éventuelles divergences avec la version 5.7. Tous les exemples sont donc réalisés avec PyQt5. Deux approches de développement vous sont proposées : la manière impérative par assemblage de composants (aussi appelés widgets) et la manière déclarative à l'aide du langage QML (Qt Quick). Une même application de gestion de bibliothèque sera développée à titre d'exemple avec chacune des méthodes. Pour aller plus loin, vous aborderez dans une troisième partie, l'affichage 2D interactif avec des vues graphiques (dans une application Qt Widget) et différents types d'affichages avancés avec Qt Quick (Canvas, Qt3D). Eric6 a été choisi comme environnement de développement par défaut. Vous utiliserez également Qt Creator pour développer en QML. [less ▲]

Detailed reference viewed: 184 (7 ULiège)
Full Text
See detailA Journey Through Julia: A dynamic and fast language
Cuvelier, Thibaut ULiege

Speech/Talk (2016)

Usually, dynamic programming languages (like Python, R, or MATLAB) are quite slow when executed, which causes performance problems in many applications. Julia is a blossoming language, both dynamic and ... [more ▼]

Usually, dynamic programming languages (like Python, R, or MATLAB) are quite slow when executed, which causes performance problems in many applications. Julia is a blossoming language, both dynamic and fast, fully open source, with a syntax that is very similar to that of MATLAB. This talk presents briefly Julia and its ecosystem. [less ▲]

Detailed reference viewed: 206 (34 ULiège)
Full Text
See detailOptimisation and uncertainty: comparing stochastic and robust programming
Cuvelier, Thibaut ULiege

Conference (2016, January 28)

Traditional optimisation tools focus on deterministic problems: scheduling airline flight crews (with as few employees as possible while still meeting legal constraints, such as maximum working time ... [more ▼]

Traditional optimisation tools focus on deterministic problems: scheduling airline flight crews (with as few employees as possible while still meeting legal constraints, such as maximum working time), finding the shortest path in a graph (used by navigation systems to give directions), etc. However, this deterministic hypothesis sometimes provides useless solutions: actual parameters cannot always be known to full precision, one reason being their randomness. For example, when scheduling trucks for freight transportation, if there is unexpected congestion on the roads, the deadlines might not be met, the company might be required to financially compensate for this delay, but also for the following deliveries that could not be made on schedule. Two main approaches are developed in the literature to take into account this uncertainty: make decision based on probability distributions of the uncertain parameters (stochastic programming) or considering they lie in a so-called ¿uncertainty set¿ (robust programming). In general, the first one leads to a large increase in the size of the problems to solve (and thus requires algorithms to work around this dimensionality curse), while the second is more conservative but tends to change the nature of the programs (which can impose a new solver technology). This talk compares the two approaches on three different cases: facility location, unit-commitment, and reservoir management. On the implementation side, multiple specific algorithms have been implemented to solve stochastic programs in order to compare their relative performance: Benders¿ decomposition, progressive hedging, and the deterministic equivalent. When comparing stochastic and robust programming, many differences appear in many aspects, even though the literature about those is very scarce. (Furthermore, those two approaches are not incompatible: both can be used in the same optimisation model to take into account different parts of the uncertainty.) Concerning solving time, stochastic programming quickly gives rise to intractable problems, which requires the development of more specific algorithm just to be able to solve them to an acceptable accuracy in decent time. What is more, the stochastic description of the uncertain values (with a discretisation of the probability distribution through scenarios) must cover all the possible uncertainty, otherwise the solution risks overfitting those scenarios, and is likely to have poor performance on close but different scenarios that may happen in practice ¿ which imposes to use a large number of scenarios, which yields very large (and hardly tractable) optimisation programs. On the other hand, by using specific uncertainty sets, robust programming yields programs that are only very slightly harder to solve, with an objective function that is very close to that of stochastic programming, but with totally different robustness properties: by using an uncertainty set computed from the scenarios, and not the scenarios themselves, it is able to withstand a much higher uncertainty than stochastic programming. However, when facing other types of uncertainty, this conclusion might turn untrue, with robust programming unable to cope with them and to bring interesting solutions to the table. [less ▲]

Detailed reference viewed: 65 (11 ULiège)
Full Text
See detailImplementing and Comparing Stochastic and Robust Programming
Cuvelier, Thibaut ULiege

Master's dissertation (2015)

Traditional optimisation tools focus on deterministic problems: scheduling airline flight crews (with as few employees as possible while still meeting legal constraints, such as maximum working time ... [more ▼]

Traditional optimisation tools focus on deterministic problems: scheduling airline flight crews (with as few employees as possible while still meeting legal constraints, such as maximum working time), finding the shortest path in a graph (used by navigation systems to give directions, usually based on GPS signals), etc. However, this deterministic hypothesis sometimes yields useless solutions: actual parameters cannot always be known to full precision, one reason being their randomness. For example, when scheduling trucks for freight transportation, if there is unexpected congestion on the roads, the deadlines might not be met, the company might be required to financially compensate for this delay, but also for the following deliveries that could not be made on schedule. Two main approaches are developed in the literature to take into account this uncertainty: take decision based on probability distributions of the uncertain parameters (stochastic programming) or considering they lie in some set (robust programming). In general, the first one leads to a large increase in the size of the problems to solve (and thus requires algorithms to work around this dimensionality curse), while the second is more conservative but tends to change the nature of the programs (which can impose a new solver technology). Some authors claim that those two mindsets are equivalent, meaning that the solutions they provide are equivalent when faced with the same uncertainty. The goal of this thesis is to explore this question: for various problems, implement those two approaches, and compare them. Is one solution more secluded from variations due to the uncertain parameters? Does it bring benefits over a deterministic approach? Is one cheaper than the other to compute? [less ▲]

Detailed reference viewed: 114 (23 ULiège)
See detailCréer des applications avec Qt 5 : les essentiels
Belz, Guillaume; Cuvelier, Thibaut ULiege; Diallo, Ilya et al

Book published by D-BookeR - 1re (2013)

Sortie fin 2012, la version 5 de Qt® marque une nouvelle étape dans la programmation multiplate-forme. Prenant en compte l'évolution de la demande en matière de développement applicatif, le framework se ... [more ▼]

Sortie fin 2012, la version 5 de Qt® marque une nouvelle étape dans la programmation multiplate-forme. Prenant en compte l'évolution de la demande en matière de développement applicatif, le framework se présente désormais comme une vaste boîte à outils modulaire pour créer toutes sortes d'applications à destination des principales plates-formes de bureau et des appareils mobiles. Créer des applications avec Qt 5 - Les essentiels vise à couvrir les outils fondamentaux de Qt 5. Que vous soyez débutant ou avancé, développeur ou designer, ce livre vous accompagnera dans la prise en main de cette nouvelle version. Au travers d'exemples riches et variés, il vous fournit toutes les clés pour développer des applications, vous aider à choisir vos modules graphiques, tirer parti de Qt Creator, optimiser votre développement, ou encore réussir votre migration depuis Qt 4. La part belle est naturellement faite à Qt Quick et au QML, qui recèlent les principales nouveautés, et c'est par leur apprentissage que devra commencer le débutant. À l'image du framework, ce livre a été conçu selon une architecture modulaire, à savoir en unités plus ou moins autonomes traitant d'un aspect particulier de Qt 5. Son objectif est de répondre à des besoins d'apprentissage et d'utilisation variés et de permettre au lecteur d'accéder directement aux parties qui l'intéressent. La plupart des modules étant disponibles à l'unité, vous pouvez choisir de n'acquérir que les parties dont vous avez besoin. Il vous présentera notamment : les nouveautés de Qt 5 les éléments de base pour démarrer avec Qt Qt Quick QGraphicsView Qt Creator [less ▲]

Detailed reference viewed: 154 (16 ULiège)
See detailWeb sémantique : méthodes et outils pour le Web de données
Heath, Tom; Bizer, Christian; Seilles, Antoine et al

Book published by Pearson Éducation France - première (2012)

Cet ouvrage est un outil de formation et de référence pour les professionnels des métiers du web (développeurs, administrateurs bases de données, Architectes) et des métiers scientifiques. Il présente l ... [more ▼]

Cet ouvrage est un outil de formation et de référence pour les professionnels des métiers du web (développeurs, administrateurs bases de données, Architectes) et des métiers scientifiques. Il présente l'ensemble des formats, techniques, méthodes et outils (RDF , SPARQL, OWL et RDFS) pour la publication d'informations sous forme de données liées sur le web, que ces données soient existantes (mais de sources ou de format différents) ou à créer. Tout au long du livre, on suit l’avancement d’un projet exemple « Production Big Lynx » au fur et à mesure de son développement. Il comporte des chapitres inédits, écrits pour l’édition française, spécifiques sur certaines notions : les ontologies (conception, définition, bonnes pratiques), les formats RDF (les différentes façons d’écrire du RDF) et les avantages de l’adoption du Web sémantique par les entreprises (référencement, accessibilité des données, faible coût). [less ▲]

Detailed reference viewed: 291 (19 ULiège)