Browsing
     by title


0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

or enter first few letters:   
OK
Full Text
See detailMachine learning techniques for atmospheric pollutant monitoring
Sainlez, Matthieu ULg; Heyen, Georges ULg

Poster (2012, January 27)

Machine learning techniques are compared to predict nitrogen oxide (NOx) pollutant emission from the recovery boiler of a Kraft pulp mill. Starting from a large database of raw process data related to a ... [more ▼]

Machine learning techniques are compared to predict nitrogen oxide (NOx) pollutant emission from the recovery boiler of a Kraft pulp mill. Starting from a large database of raw process data related to a Kraft recovery boiler, we consider a regression problem in which we are trying to predict the value of a continuous variable. Generalization is done on the worst case configuration possible to make sure the model is adequate: the training period concerns stationary operations while test periods mainly focus on NOx emissions during transient operations. [less ▲]

Detailed reference viewed: 29 (7 ULg)
Full Text
Peer Reviewed
See detailMachine learning techniques to assess the performance of a gait analysis system
Pierard, Sébastien ULg; Phan-Ba, Rémy; Van Droogenbroeck, Marc ULg

in European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN) (2014, April 24)

This paper presents a methodology based on machine learning techniques to assess the performance of a system measuring the trajectories of the lower limbs extremities for the follow-up of patients with ... [more ▼]

This paper presents a methodology based on machine learning techniques to assess the performance of a system measuring the trajectories of the lower limbs extremities for the follow-up of patients with multiple sclerosis. We show how we have established, with the help of machine learning, four important properties about this system: (1) an automated analysis of gait characteristics provides an improved analysis with respect to that of a human expert, (2) after learning, the gait characteristics provided by this system are valuable compared to measures taken by stopwatches, as used in the standardized tests, (3) the motion of the lower limbs extremities contains a lot of useful information about the gait, even if it is only a small part of the body motion, (4) a measurement system combined with a machine learning tool is sensitive to intra-subject modifications of the walking pattern. [less ▲]

Detailed reference viewed: 53 (7 ULg)
See detailMachine learning, neural networks and statistical pattern recognition for voltage security: a comparative study
Wehenkel, Louis ULg; Van Cutsem, Thierry ULg; Pavella, Mania ULg et al

in Proc. 5th International Conference on Intelligent System Applications to Power systems (1994, September)

Detailed reference viewed: 24 (0 ULg)
Peer Reviewed
See detailMachine learning, neural networks and statistical pattern recognition for voltage security: a comparative study
Wehenkel, Louis ULg; Van Cutsem, Thierry ULg; Pavella, Mania ULg et al

in International Journal of Intelligent Systems (1994), 2

Detailed reference viewed: 22 (2 ULg)
Full Text
See detailMachine learning-based feature ranking: Statistical interpretation and gene network inference
Huynh-Thu, Vân Anh ULg

Doctoral thesis (2012)

Machine learning techniques, and in particular supervised learning methods, are nowadays widely used in bioinformatics. Two prominent applications that we target specifically in this thesis are biomarker ... [more ▼]

Machine learning techniques, and in particular supervised learning methods, are nowadays widely used in bioinformatics. Two prominent applications that we target specifically in this thesis are biomarker discovery and regulatory network inference. These two problems are commonly addressed through the use of feature ranking methods that order the input features of a supervised learning problem from the most to the less relevant for predicting the output. This thesis presents, on the one hand, methodological contributions around machine learning-based feature ranking techniques and on the other hand, more applicative contributions on gene regulatory network inference. Our methodological contributions focus on the problem of selecting truly relevant features from machine learning-based feature rankings. Unlike the p-values returned by univariate tests, relevance scores derived from machine learning techniques to rank the features are usually not statistically interpretable. This lack of interpretability makes the identification of the truly relevant features among the top-ranked ones a very difficult task and hence prevents the wide adoption of these methods by practitioners. Our first contribution in this field concerns a procedure, based on permutation tests, that estimates for each subset of top-ranked features the probability for that subset to contain at least one irrelevant feature (called CER for "conditional error rate"). As a second contribution, we performed a large-scale evaluation of several, existing or novel, procedures, including our CER method, that all replace the original relevance scores with measures that can be interpreted in a statistical way. These procedures, which were assessed on several artificial and real datasets, differ greatly in terms of computing times and the tradeoff they achieve in terms of false positives and false negatives. Our experiments also clearly highlight that using model performance as a criterion for feature selection is often counter-productive. The problem of gene regulatory network inference can be formulated as several feature selection problems, each one aiming at discovering the regulators of one target gene. Within this family of methods, we developed the GENIE3 algorithm that exploits feature rankings derived from tree-based ensemble methods to infer gene networks from steady-state gene expression data. In a second step, we derived two extensions of GENIE3 that aim to infer regulatory networks from other types of data. The first extension exploits expression data provided by time course experiments, while the second extension is related to genetical genomics datasets, which contain expression data together with information about genetic markers. GENIE3 was best performer in the DREAM4 In Silico Multifactorial challenge in 2009 and in the DREAM5 Network Inference challenge in 2010, and its extensions perform very well compared to other methods on several artificial datasets. [less ▲]

Detailed reference viewed: 429 (39 ULg)
Full Text
Peer Reviewed
See detailMachine perfusion in clinical trials : "machine vs. solution effects"
Treckmann, Jürgen; Moers, Cyril; Smits, Jacqueline M et al

in Transplant International (2012), 25

Detailed reference viewed: 15 (0 ULg)
Full Text
Peer Reviewed
See detailMachine Perfusion or cold storage in deceased-donor kidney transplantation
Moers, C.; Smits, J.; Maathuis, M. H. et al

in New England Journal of Medicine [=NEJM] (2009), 360

BACKGROUND Static cold storage is generally used to preserve kidney allografts from deceased donors. Hypothermic machine perfusion may improve outcomes after transplantation, but few sufficiently powered ... [more ▼]

BACKGROUND Static cold storage is generally used to preserve kidney allografts from deceased donors. Hypothermic machine perfusion may improve outcomes after transplantation, but few sufficiently powered prospective studies have addressed this possibility. METHODS In this international randomized, controlled trial, we randomly assigned one kidney from 336 consecutive deceased donors to machine perfusion and the other to cold storage. All 672 recipients were followed for 1 year. The primary end point was delayed graft function (requiring dialysis in the first week after transplantation). Secondary end points were the duration of delayed graft function, delayed graft function defined by the rate of the decrease in the serum creatinine level, primary nonfunction, the serum creatinine level and clearance, acute rejection, toxicity of the calcineurin inhibitor, the length of hospital stay, and allograft and patient survival. RESULTS Machine perfusion significantly reduced the risk of delayed graft function. Delayed graft function developed in 70 patients in the machine-perfusion group versus 89 in the cold-storage group (adjusted odds ratio, 0.57; P = 0.01). Machine perfusion also significantly improved the rate of the decrease in the serum creatinine level and reduced the duration of delayed graft function. Machine perfusion was associated with lower serum creatinine levels during the first 2 weeks after transplantation and a reduced risk of graft failure (hazard ratio, 0.52; P = 0.03). One-year allograft survival was superior in the machine-perfusion group (94% vs. 90%, P = 0.04). No significant differences were observed for the other secondary end points. No serious adverse events were directly attributable to machine perfusion. CONCLUSIONS Hypothermic machine perfusion was associated with a reduced risk of delayed graft function and improved graft survival in the first year after transplantation. (Current Controlled Trials number, ISRCTN83876362.) [less ▲]

Detailed reference viewed: 33 (7 ULg)
Full Text
Peer Reviewed
See detailMachine perfusion versus cold storage for preservation of kidneys from expanded criteria donors after brain death
Treckmann, Jürgen; Moers, Cyril; Smits, Jacqueline M et al

in Transplant International (2011), 24

The purpose of this study was to analyze the possible effects of machine perfusion (MP) versus cold storage (CS) on delayed graft function (DGF) and early graft survival in expanded criteria donor kidneys ... [more ▼]

The purpose of this study was to analyze the possible effects of machine perfusion (MP) versus cold storage (CS) on delayed graft function (DGF) and early graft survival in expanded criteria donor kidneys (ECD). As part of the previously reported international randomized controlled trial 91 consecutive heartbeating deceased ECDs – defined according to the United Network of Organ Sharing definition – were included in the study. From each donor one kidney was randomized to MP and the contralateral kidney to CS. All recipients were followed for 1 year. The primary endpoint was DGF. Secondary endpoints included primary nonfunction and graft survival. DGF occurred in 27 patients in the CS group (29.7%) and in 20 patients in the MP group (22%). Using the logistic regression model MP significantly reduced the risk of DGF compared with CS (OR 0.460, P = 0.047). The incidence of nonfunction in the CS group (12%) was four times higher than in the MP group (3%) (P = 0.04). One-year graft survival was significantly higher in machine perfused kidneys compared with cold stored kidneys (92.3% vs. 80.2%, P = 0.02). In the present study, MP preservation clearly reduced the risk of DGF and improved 1-year graft survival and function in ECD kidneys. [less ▲]

Detailed reference viewed: 18 (2 ULg)
Full Text
Peer Reviewed
See detailMachine perfusion versus cold storage for the preservation of kidneys from donors ≥ years allocated in the Eurotransplant Senior Programme
GALLINAT, Anja; MOERS, Cyril; TRECKMANN, Jürgen et al

in Nephrology Dialysis Transplantation (2012), 27

Background. In the Eurotransplant Senior Programme (ESP), kidneys from donors aged ≥65 years are preferentially allocated locally and transplanted into patients aged ≥65 years on dialysis. The purpose of ... [more ▼]

Background. In the Eurotransplant Senior Programme (ESP), kidneys from donors aged ≥65 years are preferentially allocated locally and transplanted into patients aged ≥65 years on dialysis. The purpose of this study was to analyse whether the results of transplantation in the ESP can be improved by preservation of organs by hypothermic machine perfusion (MP) compared with simple cold storage (CS). Methods. Overall, 85 deceased heart-beating donors ≥65 years of age were included in this analysis with follow-up until 1 year post-transplant. For each donor, one kidney was randomly assigned to preservation by CS and the contralateral kidney to MP from organ procurement until transplantation. Delayed graft function (DGF), primary non-function (PNF) and 1-year patient and graft survival rates were evaluated as primary and secondary endpoints. Results. The median recipient age was 66 years in both groups and the median cold ischaemia time was 11 h for MP and 10.5 h for CS (P = 0.69). The DGF rate was 29.4% for MP and 34.1% for CS (P = 0.58). Only extended duration of cold ischaemia time was an independent risk factor for the development of DGF (odds ratio 1.2, P < 0.0001). PNF was significantly reduced (3.5% MP versus 12.9% CS, P = 0.02). The 1-year patient and graft survival rates were similar for MP and CS (94% versus 95% and 89 versus 81%, P > 0.05). The 1-year graft survival rate was significantly improved after MP in recipients who developed DGF (84% MP versus 48% CS, P = 0.01). Conclusions. Continuous pulsatile hypothermic MP for kidneys from donors aged ≥65 years can reduce the rate of never-functioning kidneys and improve the 1-year graft survival rate of kidneys with DGF. In this small cohort, the known advantage of MP for the reduction of DGF could not be confirmed, possibly due to relatively short cold ischaemia times. [less ▲]

Detailed reference viewed: 25 (2 ULg)
Peer Reviewed
See detailA Machine-Intelligent System for Automatic Target Recognition
Dudgeon, Dan E.; Verly, Jacques ULg; Delanoy, Richard L.

Conference (1990, November)

Detailed reference viewed: 3 (0 ULg)
Full Text
Peer Reviewed
See detailMachine-learning approaches to power-system security assessment
Wehenkel, Louis ULg

in IEEE Expert (1997), 12(5), 60-72

The paper discusses a framework that uses machine learning and other automatic-learning methods to assess power-system security. The framework exploits simulation models in parallel to screen diverse ... [more ▼]

The paper discusses a framework that uses machine learning and other automatic-learning methods to assess power-system security. The framework exploits simulation models in parallel to screen diverse simulation scenarios of a system, yielding a large database. Using data mining techniques, the framework extracts synthetic information about the simulated system's main features from this database [less ▲]

Detailed reference viewed: 14 (2 ULg)
Full Text
Peer Reviewed
See detailMachine-learnt versus analytical models of TCP throughput
El Khayat, Ibtissam; Geurts, Pierre ULg; Leduc, Guy ULg

in Computer Networks (2007), 51(10), 2631-2644

We first study the accuracy of two well-known analytical models of the average throughput of long-term TCP flows, namely the so-called SQRT and PFTK models, and show that these models are far from being ... [more ▼]

We first study the accuracy of two well-known analytical models of the average throughput of long-term TCP flows, namely the so-called SQRT and PFTK models, and show that these models are far from being accurate in general. Our simulations, based on a large set of long-term TCP sessions, show that 70% of their predictions exceed the boundaries of TCP-Friendliness, thus questioning their use in the design of new TCP-Friendly transport protocols. We then investigate the reasons of this inaccuracy, and show that it is largely due to the lack of discrimination between the two packet loss detection methods used by TCP, namely by triple duplicate acknowledgements or by timeout: expirations. We then apply various machine learning techniques to infer new models of the average TCP throughput. We show that they are more accurate than the SQRT and PFTK models, even without the above discrimination, and are further improved when we allow the machine-learnt models to distinguish the two loss detection techniques. Although our models are not analytical formulas, they can be plugged in transport protocols to make them TCP-Friendly. Our results also suggest that analytical models of the TCP throughput should certainly benefit from the incorporation of the timeout loss rate. (C) 2006 Elsevier B.V. All rights reserved. [less ▲]

Detailed reference viewed: 35 (3 ULg)
Full Text
See detailMachines à concepts
Durand, Pascal ULg

Article for general public (1996)

Detailed reference viewed: 7 (1 ULg)
Full Text
See detailMachines et Systèmes Thermiques
Lebrun, Jean ULg; Lemort, Vincent ULg

Learning material (2007)

Detailed reference viewed: 120 (17 ULg)
Full Text
Peer Reviewed
See detailMachining process simulation using Samcef superelement
Masset, Luc ULg; Debongnie, Jean-François ULg; Marty, Audrey

Conference (2005, February)

In this paper, we present a new simulation tool for process engineers. During process design phases, several aspects of machining have to be taken into account. Classical CAD/CAM suites still lack some ... [more ▼]

In this paper, we present a new simulation tool for process engineers. During process design phases, several aspects of machining have to be taken into account. Classical CAD/CAM suites still lack some crucial issues. The goal of the developed tool is to predict the geometric errors of machined surfaces. For classical applications of the automotive domain, form errors are mainly due to the machined part and clamping system flexibility. They are modeled thanks to the FE method. The major peculiarity of the adopted model is to apply numerous load cases. To achieve a low computational coast, we have combined the SAMCEF superelement feature and a specific code to solve the reduced system. This original scheme allows solving efficiently large industrial applications. [less ▲]

Detailed reference viewed: 102 (6 ULg)
Full Text
Peer Reviewed
See detailMachining processes simulation : specific finite element aspects
Masset, Luc ULg; Debongnie, Jean-François ULg

in Journal of Computational & Applied Mathematics (2004, July 01), 168(1-2), 309-320

The paper presents a simulation tool designed to predict form errors of part surfaces obtained by face milling and turning processes. For these operations, the form error is often due to the flexibility ... [more ▼]

The paper presents a simulation tool designed to predict form errors of part surfaces obtained by face milling and turning processes. For these operations, the form error is often due to the flexibility of the workpiece and its supports. The finite element method is adopted to model the part geometry and to compute its deformations. Numerous load cases are required to obtain the form error so that classical resolution methods prove to be inefficient (CPU time, memory and disk space). The paper mainly focuses on the special computation scheme adopted in order to improve the resolution of such an atypical problem. [less ▲]

Detailed reference viewed: 85 (31 ULg)
Full Text
See detailMacro Minitab pour l'analyse discriminante logistique
Palm, Rodolphe ULg

Software (2008)

Detailed reference viewed: 14 (2 ULg)