References of "Wehenkel, Louis"
     in
Bookmark and Share    
Full Text
Peer Reviewed
See detailPost-contingency corrective control failure: a risk to neglect or a risk to control?
Karangelos, Efthymios ULiege; Wehenkel, Louis ULiege

Conference (in press)

This paper proposes a methodology for assessing the risk implied by the potential failure of post-contingency corrective controls. We express such risk in terms of service interruption socio-economic ... [more ▼]

This paper proposes a methodology for assessing the risk implied by the potential failure of post-contingency corrective controls. We express such risk in terms of service interruption socio-economic severity to the system end-consumers and argue for considering its magnitude not only in absolute terms, but most importantly in relation to a spectrum of socio-economic metrics fully describing the operation of an electrical power system as per the applicable reliability management approach (presently based on the N-1 criterion). We showcase the proposed methodology by presenting its application through case studies on the single area version of the IEEE-RTS96 benchmark. Our analysis establishes that the proposed assessment scope is quite informative in distinguishing whether the risk implied by the potential failure of post-contingency corrective control is noteworthy or negligible. [less ▲]

Detailed reference viewed: 44 (4 ULiège)
Full Text
See detailIntegrating facial detection and recognition algorithms into real-life applications
Van Lishout, François ULiege; Dubois, Antoine ULiege; Wang, Menglan Linda ULiege et al

Conference (2018, June 27)

Performances of facial detection and recognition algorithms on publicly available datasets do not always reflect their true effectiveness in practical real-life applications. Parameters such as distance ... [more ▼]

Performances of facial detection and recognition algorithms on publicly available datasets do not always reflect their true effectiveness in practical real-life applications. Parameters such as distance to camera, blur or pose, which vary across datasets, have an important impact on performances. Furthermore, computing speed may also be a key factor for applications requiring real-time decisions. In our department, we work on an application localizing any registered user present in the building in real-time (we also provide an application allowing users to manage their privacy), based only on a few pictures automatically taken during the registration process. In this work, we first compare four open-source facial detection algorithms on the WIDER FACE dataset and on an independent one constructed in our department with volunteers, containing images having a large variation in terms of size, pose, illumination and level of blur. We show that Single Stage Headless Face Detector (SSH) leads to way better precision- recall performances, but is about twice slower than the second best method Faster R-CNN. Second, we compare three open-source facial recognition algorithms on the MegaFace dataset and on our above mentioned one. The latter shows to be much more challenging for all methods, suggesting that publications comparing methods on the former may display performances that cannot be achieved in real-life contexts. We show that InsightFace leads to slightly better precision-recall performances than Dlib, but is about three time slower than the latter. [less ▲]

Detailed reference viewed: 24 (0 ULiège)
Full Text
Peer Reviewed
See detailRandom Subspace with Trees for Feature Selection Under Memory Constraints
Sutera, Antonio ULiege; Châtel, Célia; Louppe, Gilles ULiege et al

in Storkey, Amos; Perez-Cruz, Fernando (Eds.) Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics (2018)

Dealing with datasets of very high dimension is a major challenge in machine learning. In this paper, we consider the problem of feature selection in applications where the memory is not large enough to ... [more ▼]

Dealing with datasets of very high dimension is a major challenge in machine learning. In this paper, we consider the problem of feature selection in applications where the memory is not large enough to contain all features. In this setting, we propose a novel tree-based feature selection approach that builds a sequence of randomized trees on small subsamples of variables mixing both variables already identified as relevant by previous models and variables randomly selected among the other variables. As our main contribution, we provide an in-depth theoretical analysis of this method in infinite sample setting. In particular, we study its soundness with respect to common definitions of feature relevance and its convergence speed under various variable dependance scenarios. We also provide some preliminary empirical results highlighting the potential of the approach. [less ▲]

Detailed reference viewed: 19 (2 ULiège)
Full Text
Peer Reviewed
See detailUsing Machine Learning to Enable Probabilistic Reliability Assessment in Operation Planning
Duchesne, Laurine ULiege; Karangelos, Efthymios ULiege; Wehenkel, Louis ULiege

in Power Systems Computation Conference 2018 Proceedings (2018)

In the context of operation planning, probabilistic reliability assessment essentially boils down to predicting, efficiently and with sufficient accuracy, various economic and reliability indicators ... [more ▼]

In the context of operation planning, probabilistic reliability assessment essentially boils down to predicting, efficiently and with sufficient accuracy, various economic and reliability indicators reflecting the expected performance of the system over a certain look-ahead horizon, so as to guide the operation planner in his decision-making. In order to speed-up the crude Monte Carlo approach, which would entail a very large number of heavy computations, we propose in this paper an approach combining Monte Carlo simulation, machine learning and variance reduction techniques such as control variates. We provide an extensive case study testing this approach on the three-area IEEE-RTS96 benchmark, in the context of day-ahead operation planning while using a security constrained optimal power flow model to simulate real-time operation according to the N-1 criterion. From this case study, we can conclude that the proposed approach allows to reduce the number of heavy computations by about an order of magnitude, without sacrificing accuracy. [less ▲]

Detailed reference viewed: 76 (9 ULiège)
Full Text
Peer Reviewed
See detailProbabilistic Reliability Management Approach and Criteria for Power System Short-term Operational Planning
Karangelos, Efthymios ULiege; Wehenkel, Louis ULiege

in Probabilistic Reliability Management Approach and Criteria for Power System Short-term Operational Planning (2017, August)

This paper develops a probabilistic decision making framework for reliability management in the short-term operational planning context. We build upon our recent work, which proposed a probabilistic ... [more ▼]

This paper develops a probabilistic decision making framework for reliability management in the short-term operational planning context. We build upon our recent work, which proposed a probabilistic reliability management approach and criterion (RMAC) for the latest decision making opportunity of real-time system operation. Here, we transpose the RMAC to the preceding problem instance of short-term operational planning, wherein i) risk is aggravated by the uncertainty on power injections and weather conditions, and, ii) the problem scope concerns choosing `strategic' actions (e.g., starting additional generating units, granting outage requests for maintenance, etc.) to facilitate decision making during the forthcoming real-time system operation. To anticipate on the latter, we formalize the notion of a real-time `proxy' as a simplified model of the real-time decision making context, adequately accurate for the purpose of operational planning decision making. Stating a first proposal for such a proxy, we mathematically formulate the RMAC for short-term operational planning as a multi-stage stochastic decision making problem and demonstrate its main features by case studies on a modified version of the single area IEEE RTS-96 system. [less ▲]

Detailed reference viewed: 84 (12 ULiège)
Full Text
Peer Reviewed
See detailA Machine Learning-Based Approximation of Strong Branching
Marcos Alvarez, Alejandro ULiege; Louveaux, Quentin ULiege; Wehenkel, Louis ULiege

in INFORMS Journal on Computing (2017), 29(1), 185-195

We present in this paper a new generic approach to variable branching in branch-and-bound for mixed- integer linear problems. Our approach consists in imitating the decisions taken by a good branching ... [more ▼]

We present in this paper a new generic approach to variable branching in branch-and-bound for mixed- integer linear problems. Our approach consists in imitating the decisions taken by a good branching strategy, namely strong branching, with a fast approximation. This approximated function is created by a machine learning technique from a set of observed branching decisions taken by strong branching. The philosophy of the approach is similar to reliability branching. However, our approach can catch more complex aspects of observed previous branchings in order to take a branching decision. The experiments performed on randomly generated and MIPLIB problems show promising results. [less ▲]

Detailed reference viewed: 234 (17 ULiège)
Full Text
Peer Reviewed
See detailA computational model of mid-term outage scheduling for long-term system studies
Marin, Manuel ULiege; Karangelos, Efthymios ULiege; Wehenkel, Louis ULiege

in PowerTech Manchester 2017 Proceedings (2017)

This paper presents a computational model of the mid-term outage scheduling process of electric power transmis- sion assets, to be used in long-term studies such as mainte- nance policy assessments and ... [more ▼]

This paper presents a computational model of the mid-term outage scheduling process of electric power transmis- sion assets, to be used in long-term studies such as mainte- nance policy assessments and system development studies, while accounting for the impact of outage schedules on short-term system operation. We propose a greedy algorithm that schedules the outages one by one according to their impact on system operation estimated via Monte-Carlo simulations. The model is implemented in JULIA and applied to the IEEE RTS-96. [less ▲]

Detailed reference viewed: 53 (9 ULiège)
Full Text
Peer Reviewed
See detailMachine Learning of Real-time Power Systems Reliability Management Response
Duchesne, Laurine ULiege; Karangelos, Efthymios ULiege; Wehenkel, Louis ULiege

in PowerTech Manchester 2017 Proceedings (2017)

In this paper we study how supervised machine learning could be applied to build simplified models of real-time (RT) reliability management response to the realization of uncertainties. The final ... [more ▼]

In this paper we study how supervised machine learning could be applied to build simplified models of real-time (RT) reliability management response to the realization of uncertainties. The final objective is to import these models into look-ahead operation planning under uncertainties. Our response models predict in particular the real-time reliability management costs and the resulting reliability level of the system. We tested our methodology on the IEEE-RTS96 benchmark. Among the supervised learning algorithms tested, extremely randomized trees, kernel ridge regression and neural networks appear to be the best methods for this application. Furthermore, by using feature “importances” computed by tree-based ensemble methods, we were able to extract the most relevant variables to predict the response of real-time reliability management, and thus obtain a better understanding of the system properties. [less ▲]

Detailed reference viewed: 240 (41 ULiège)
Full Text
See detailBig data, machine learning, and optimization, for power systems reliability
Wehenkel, Louis ULiege

Scientific conference (2016, November 09)

How to combine physical models with observational data for ensuring power systems reliability, by leveraging simulation, optimisation, and machine learning.

Detailed reference viewed: 190 (16 ULiège)
Full Text
Peer Reviewed
See detailRandom subspace with trees for feature selection under memory constraints
Sutera, Antonio ULiege; Châtel, Célia; Louppe, Gilles ULiege et al

Conference (2016, September 12)

Detailed reference viewed: 248 (28 ULiège)
Full Text
Peer Reviewed
See detailAutomatic learning of fine operating rules for online power system security control
Sun, Hongbin; Zhao, Feng; Wang, Huifang et al

in IEEE Transactions on Neural Networks and Learning Systems (2016), 27(8), 1708-1719

Fine operating rules for security control and an automatic system for their online discovery were developed to adapt to the development of smart grids. The automatic system uses the real-time system state ... [more ▼]

Fine operating rules for security control and an automatic system for their online discovery were developed to adapt to the development of smart grids. The automatic system uses the real-time system state to determine critical flowgates, and then a continuation power flow-based security analysis is used to compute the initial transfer capability of critical flowgates. Next, the system applies the Monte Carlo simulations to expected short-term operating condition changes, feature selection, and a linear least squares fitting of the fine operating rules. The proposed system was validated both on an academic test system and on a provincial power system in China. The results indicated that the derived rules provide accuracy and good interpretability and are suitable for real-time power system security control. The use of high-performance computing systems enables these fine operating rules to be refreshed online every 15 min. [less ▲]

Detailed reference viewed: 60 (5 ULiège)
Full Text
Peer Reviewed
See detailMettre en place des « tableaux » de bord dans l’étude de l’histologie – Une exploration du potentiel pédagogique des traces d’apprentissage
Verpoorten, Dominique ULiege; Vincke, Grégoire ULiege; Pesesse, Laurence ULiege et al

Conference (2016, June 06)

La communication synthétise les résultats d'un questionnaire visant à estimer les attentes d'enseignants en histologie en matière de visualisations de traces d'apprentissage laissées par leurs étudiants ... [more ▼]

La communication synthétise les résultats d'un questionnaire visant à estimer les attentes d'enseignants en histologie en matière de visualisations de traces d'apprentissage laissées par leurs étudiants lorsqu'ils travaillent avec un outil spécialisé (Cytomine) permettant une interaction avec des coupes histologiques numérisées. [less ▲]

Detailed reference viewed: 63 (23 ULiège)
Full Text
Peer Reviewed
See detailContext-dependent feature analysis with random forests
Sutera, Antonio ULiege; Louppe, Gilles ULiege; Huynh-Thu, Vân Anh ULiege et al

in Uncertainty In Artificial Intelligence: Proceedings of the Thirty-Two Conference (2016) (2016, June)

Detailed reference viewed: 145 (36 ULiège)
Full Text
Peer Reviewed
See detailProbabilistic reliability management approach and criteria for power system real-time operation
Karangelos, Efthymios ULiege; Wehenkel, Louis ULiege

in Power Systems Computation Conference (2016, June)

This paper develops a probabilistic approach for power system reliability management in real-time operation where risk is a product of i) the potential occurrence of contingencies, ii) the possible ... [more ▼]

This paper develops a probabilistic approach for power system reliability management in real-time operation where risk is a product of i) the potential occurrence of contingencies, ii) the possible failure of corrective (i.e., post-contingency) control and, iii) the socio-economic impact of service interruptions to end-users. Stressing the spatiotemporal variability of these factors, we argue for reliability criteria assuring a high enough probability of avoiding service interruptions of severe socio-economic impact by dynamically identifying events of nonnegligible implied risk. We formalise the corresponding decision making problem as a chance-constrained two-stage stochastic programming problem, and study its main features on the single area IEEE RTS-96 system. We also discuss how to leverage this proposal for the construction of a globally coherent reliability management framework for long-term system development, midterm asset management, and short-term operation planning. [less ▲]

Detailed reference viewed: 399 (36 ULiège)
Full Text
See detailComments on: A random forest guided tour
Geurts, Pierre ULiege; Wehenkel, Louis ULiege

in TEST (2016), 25(2), 247-253

Detailed reference viewed: 28 (2 ULiège)
Full Text
Peer Reviewed
See detailFramework for Threat Based Failure Rates in Transmission System Operation
Perkin, Samuel; Bjornsson, Gudjon; Baldursdottir, Iris et al

in Framework for Threat Based Failure Rates in Transmission System Operation (2016, February)

Reliability of electrical transmission systems isvpresently managed by applying the deterministic N-1 criterion, or some variant thereof. This means that transmission systems are designed with at least ... [more ▼]

Reliability of electrical transmission systems isvpresently managed by applying the deterministic N-1 criterion, or some variant thereof. This means that transmission systems are designed with at least one level of redundancy, regardless of the cost of doing so, or the severity of the risks they mitigate. In an operational context, the N-1 criterion provides a reliability target but it fails to accurately capture the dynamic nature of short-term threats to transmission systems. Ongoing research aims to overcome this shortcoming by proposing new probabilistic reliability criteria. Such new criteria are anticipated to rely heavily on component failure rate calculations. This paper provides a threat modelling framework, using the Icelandic transmission system as an example, highlighting the need for improved data collection and failure rate modelling. The feasibility of using threat credibility indicators to achieve spatio-temporal failure rates, given minimal data, is explored in a case study of the Icelandic transmission system. The paper closes with a discussion on the assumptions and simplifications that are implicitly made in the formulation, and the additional work required for such an approach to be included in existing practices. Specifically, this paper is concerned only with short term and real-time management of electrical transmission systems. [less ▲]

Detailed reference viewed: 162 (8 ULiège)
Full Text
Peer Reviewed
See detailCollaborative analysis of multi-gigapixel imaging data using Cytomine
Marée, Raphaël ULiege; Rollus, Loïc; Stévens, Benjamin et al

in Bioinformatics (2016)

Motivation: Collaborative analysis of massive imaging datasets is essential to enable scientific discoveries. Results: We developed Cytomine to foster active and distributed collaboration of ... [more ▼]

Motivation: Collaborative analysis of massive imaging datasets is essential to enable scientific discoveries. Results: We developed Cytomine to foster active and distributed collaboration of multidisciplinary teams for large-scale image-based studies. It uses web development methodologies and machine learning in order to readily organize, explore, share, and analyze (semantically and quantitatively) multi-gigapixel imaging data over the internet. We illustrate how it has been used in several biomedical applications. Availability: Cytomine (http://www.cytomine.be/) is freely available under an open-source license from http://github.com/cytomine/. A documentation wiki (http://doc.cytomine.be) and a demo server (http://demo.cytomine.be) are also available. [less ▲]

Detailed reference viewed: 166 (36 ULiège)
Full Text
Peer Reviewed
See detailTowards Generic Image Classification using Tree-based Learning: an Extensive Empirical Study
Marée, Raphaël ULiege; Geurts, Pierre ULiege; Wehenkel, Louis ULiege

in Pattern Recognition Letters (2016), 74(15), 17-23

This paper considers the general problem of image classification without using any prior knowledge about image classes. We study variants of a method based on supervised learning whose common steps are ... [more ▼]

This paper considers the general problem of image classification without using any prior knowledge about image classes. We study variants of a method based on supervised learning whose common steps are the extraction of random subwindows described by raw pixel intensity values and the use of ensemble of extremely randomized trees to directly classify images or to learn image features. The influence of method parameters and variants is thoroughly evaluated so as to provide baselines and guidelines for future studies. Detailed results are provided on 80 publicly available datasets that depict very diverse types of images (more than 3800 image classes and over 1.5 million images). [less ▲]

Detailed reference viewed: 382 (68 ULiège)
Full Text
See detailOnline Learning for Strong Branching Approximation in Branch-and-Bound
Marcos Alvarez, Alejandro ULiege; Wehenkel, Louis ULiege; Louveaux, Quentin ULiege

E-print/Working paper (2016)

We present an online learning approach to variable branching in branch-and-bound for mixed-integer linear problems. Our approach consists in learning strong branching scores in an online fashion and in ... [more ▼]

We present an online learning approach to variable branching in branch-and-bound for mixed-integer linear problems. Our approach consists in learning strong branching scores in an online fashion and in using them to take branching decisions. More specifically, numerical scores are used to rank the branching candidates. If, for a given variable, the learned approximation is deemed reliable, then the score for that variable is computed thanks to the learned function. If the approximation is not reliable yet, the real strong branching score is used instead. The scores that are computed through the real strong branching procedure are fed to the online learning algorithm in order to improve the approximated function. The experiments show promising results. [less ▲]

Detailed reference viewed: 133 (4 ULiège)
Full Text
Peer Reviewed
See detailgammaMAXT: a fast multiple-testing correction algorithm
Van Lishout, François ULiege; Gadaleta, Francesco; Moore, Jason H. et al

in BioData Mining (2015), 8(36),

Background: The purpose of the maxT algorithm is to provide a significance test algorithm that controls the family-wise error rate (FWER) during simultaneous hypothesis testing. However, the requirements ... [more ▼]

Background: The purpose of the maxT algorithm is to provide a significance test algorithm that controls the family-wise error rate (FWER) during simultaneous hypothesis testing. However, the requirements in terms of computing time and memory of this procedure are proportional to the number of investigated hypotheses. The memory issue has been solved in 2013 by Van Lishout’s implementation of MaxT, which makes the memory usage independent from the size of the dataset. This algorithm is implemented in MBMDR-3.0.3, a software that is able to identify genetic interactions, for a variety of SNP-SNP based epistasis models effectively. On the other hand, that implementation turned out to be less suitable for genome-wide interaction analysis studies, due to the prohibitive computational burden. Results: In this work we introduce gammaMAXT, a novel implementation of the maxT algorithm for multiple testing correction. The algorithm was implemented in software MBMDR-4.2.2, as part of the MB-MDR framework to screen for SNP-SNP, SNP-environment or SNP-SNP-environment interactions at a genome-wide level. We show that, in the absence of interaction effects, test-statistics produced by the MB-MDR methodology follow a mixture distribution with a point mass at zero and a shifted gamma distribution for the top 10 % of the strictly positive values. We show that the gammaMAXT algorithm has a power comparable to MaxT and maintains FWER, but requires less computational resources and time. We analyze a dataset composed of 106 SNPs and 1000 individuals within one day on a 256-core computer cluster. The same analysis would take about 104 times longer with MBMDR-3.0.3. Conclusions: These results are promising for future GWAIs.However, the proposed gammaMAXT algorithm offers a general significance assessment and multiple testing approach, applicable to any context that requires performing hundreds of thousands of tests. It offers new perspectives for fast and efficient permutation-based significance assessment in large-scale (integrated) omics studies. [less ▲]

Detailed reference viewed: 48 (11 ULiège)