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
Peer Reviewed
See detailOptimized look-ahead tree policies: a bridge between look-ahead tree policies and direct policy search
Jung, Tobias ULg; Wehenkel, Louis ULg; Ernst, Damien ULg et al

in International Journal of Adaptive Control and Signal Processing (2014), 28(3-5), 255-289

Direct policy search (DPS) and look-ahead tree (LT) policies are two popular techniques for solving difficult sequential decision-making problems. They both are simple to implement, widely applicable ... [more ▼]

Direct policy search (DPS) and look-ahead tree (LT) policies are two popular techniques for solving difficult sequential decision-making problems. They both are simple to implement, widely applicable without making strong assumptions on the structure of the problem, and capable of producing high performance control policies. However, computationally both of them are, each in their own way, very expensive. DPS can require huge offline resources (effort required to obtain the policy) to first select an appropriate space of parameterized policies that works well for the targeted problem, and then to determine the best values of the parameters via global optimization. LT policies do not require any offline resources; however, they typically require huge online resources (effort required to calculate the best decision at each step) in order to grow trees of sufficient depth. In this paper, we propose optimized look-ahead trees (OLT), a model-based policy learning scheme that lies at the intersection of DPS and LT. In OLT, the control policy is represented indirectly through an algorithm that at each decision step develops, as in LT using a model of the dynamics, a small look-ahead tree until a prespecified online budget is exhausted. Unlike LT, the development of the tree is not driven by a generic heuristic; rather, the heuristic is optimized for the target problem and implemented as a parameterized node scoring function learned offline via DPS. We experimentally compare OLT with pure DPS and pure LT variants on optimal control benchmark domains. The results show that the LT-based representation is a versatile way of compactly representing policies in a DPS scheme (which results in OLT being easier to tune and having lower offline complexity than pure DPS); while at the same time, DPS helps to significantly reduce the size of the look-ahead trees that are required to take high-quality decisions (which results in OLT having lower online complexity than pure LT). Moreover, OLT produces overall better performing policies than pure DPS and pure LT and also results in policies that are robust with respect to perturbations of the initial conditions. [less ▲]

Detailed reference viewed: 94 (32 ULg)
Full Text
Peer Reviewed
See detailOptimized Look-Ahead Trees: Extensions to Large and Continuous Action Spaces
Jung, Tobias ULg; Ernst, Damien ULg; Maes, Francis

in Proc. of IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning (ADPRL'13) (2013)

This paper studies look-ahead tree based control policies from the viewpoint of online decision making with constraints on the computational budget allowed per decision (expressed as number of calls to ... [more ▼]

This paper studies look-ahead tree based control policies from the viewpoint of online decision making with constraints on the computational budget allowed per decision (expressed as number of calls to the generative model). We consider optimized look-ahead tree (OLT) policies, a recently introduced family of hybrid techniques, which combine the advantages of look-ahead trees (high precision) with the advantages of direct policy search (low online cost) and which are specifically designed for limited online budgets. We present two extensions of the basic OLT algorithm that on the one side allow tackling deterministic optimal control problems with large and continuous action spaces and that on the other side can also help to further reduce the online complexity. [less ▲]

Detailed reference viewed: 15 (7 ULg)
Full Text
Peer Reviewed
See detailOptimized Remeshing for Incremental Forming Simulation
Rabahallah, François ULg; Bouffioux, Chantal ULg; Duchene, Laurent ULg et al

in Proceedings of the Advances in Materials and Processing Technologies Conference (2010)

This work presents an advanced numerical methodology which aims to improve the simulations of incremental forming processes [3]. A specific numerical treatment is required to optimize the remeshing ... [more ▼]

This work presents an advanced numerical methodology which aims to improve the simulations of incremental forming processes [3]. A specific numerical treatment is required to optimize the remeshing procedure, in order to decrease the simulation CPU time and deals with the strong process nonlinearities. The algorithm remeshing procedure is implemented in the Lagamine FE code [2] developed at the University of Liège. An elastoplastic law is combined with an isotropic kinematic hardening model. The identification procedure for the material parameters determination takes into account a set of mechanical tests (inverse approach). Numerical simulations of the forming of a pyramid part with two wall angles are performed in order to address the robustness of the implementation and the ability of this numerical methodology to predict the force values. Comparisons between measured and predicted force values are performed. [less ▲]

Detailed reference viewed: 92 (16 ULg)
Full Text
Peer Reviewed
See detailOptimized Schwarz Algorithm with Double Sweep Preconditioner for the Helmholtz Equation
Vion, Alexandre ULg; Geuzaine, Christophe ULg

in Proceedings of the 9th International Symposium on Electric and Magnetic Fields, EMF 2013 (2013, April)

Detailed reference viewed: 6 (0 ULg)
Full Text
Peer Reviewed
See detailOptimized thermal zone controller for integration within a building energy management system.
Kummert, M.; Andre, Philippe ULg; Nicolas, Jacques ULg

in Proceedings Clima 2000 (1997)

Detailed reference viewed: 23 (1 ULg)
Peer Reviewed
See detailOptimized Waveform Relaxation Methods for Modeling Electromagnetic Field-Circuit Problems
Nshimiyimana, Jean de Dieu ULg; Plumier, Frédéric ULg; Dular, Patrick ULg et al

in Proceedings of Sixteenth Biennal IEEE Conference on Electromagnetic Field Computation (2014, April)

The convergence of classical and optimized waveform relaxation methods is compared for the simulation of field-circuit problems, where a finite element solver is used for the solution of the field problem ... [more ▼]

The convergence of classical and optimized waveform relaxation methods is compared for the simulation of field-circuit problems, where a finite element solver is used for the solution of the field problem and a circuit simulator is used for the electronic circuit. [less ▲]

Detailed reference viewed: 23 (8 ULg)
Full Text
Peer Reviewed
See detailOptimizing active avoidance conditioning for high-throughput behavioral and cognitive screening in fish. Proceedings of Measuring behaviour
Ylieff, Marc ULg; Ferrara, André ULg; Fagniart, C. et al

in A.J. Spink et al. (Ed.) Proceedings of Measuring Behaviour (2008)

Detailed reference viewed: 100 (40 ULg)
Full Text
Peer Reviewed
See detailOptimizing early nutritional support based on recent recommendations in VLBW infants and postnatal growth restriction
SENTERRE, Thibault ULg; Rigo, Jacques ULg

in Journal of Pediatric Gastroenterology and Nutrition (2011), 53(5), 536-542

Detailed reference viewed: 19 (1 ULg)
Peer Reviewed
See detailOptimizing Early Nutritional Support Based on Recent Recommendations in VLBW Infants.
SENTERRE, Thibault ULg

in 4th Congress of the European Academy of Paediatric Societies (2012, October)

Detailed reference viewed: 15 (2 ULg)
Full Text
Peer Reviewed
See detailOptimizing nutrition after birth with a unique standardized parenteral nutrition solution may reduce electrolytes anomalies in <1250g infants
SENTERRE, Thibault ULg; rigo, jacques

in Archives of Disease in Childhood (2012), 97(S2), 394

Detailed reference viewed: 7 (1 ULg)
Peer Reviewed
See detailOptimizing Nutritional Supply Reduces Postnatal Growth Restriction in VLBW Infants
SENTERRE, Thibault ULg; Habibi, Fakher; Rigo, J

in Pediatric Academic Societies and Asian Society for Pediatric Research Joint Meeting (2011)

Detailed reference viewed: 10 (0 ULg)
Full Text
Peer Reviewed
See detailOptimizing patient adherence and persistence in strontium ranelate phase 3 program using a short term run-in study
Reginster, Jean-Yves ULg; Spector, Tim; Badurski, J. et al

in Osteoporosis International (2002, November), 13(Suppl.3), 35-36

Detailed reference viewed: 13 (2 ULg)
Full Text
Peer Reviewed
See detailOptimizing Perfectly Matched Layers in Discrete Contexts
Modave, Axel ULg; Delhez, Eric ULg; Geuzaine, Christophe ULg

in International Journal for Numerical Methods in Engineering (2014), 99(6), 410437

Perfectly Matched Layers (PMLs) are widely used for the numerical simulation of wave-like problems defined on large or infinite spatial domains. However, for both the time-dependent and the time-harmonic ... [more ▼]

Perfectly Matched Layers (PMLs) are widely used for the numerical simulation of wave-like problems defined on large or infinite spatial domains. However, for both the time-dependent and the time-harmonic cases, their performance critically depends on the so-called absorption function. This paper deals with the choice of this function when classical numerical methods are used (based on finite differences, finite volumes, continuous finite elements and discontinuous finite elements). After reviewing the properties of the PMLs at the continuous level, we analyse how they are altered by the different spatial discretizations. In the light of these results, different shapes of absorption function are optimized and compared by means of both one- and two-dimensional representative time-dependent cases. This study highlights the advantages of the so-called shifted hyperbolic function, which is efficient in all cases and does not require the tuning of a free parameter, by contrast with the widely used polynomial functions. [less ▲]

Detailed reference viewed: 44 (10 ULg)
Full Text
Peer Reviewed
See detailOptimizing red sorghum malt quality when Bacillus subtilis is used during steeping to control mold growth
Bwanganga Tawaba, Jean-Claude ULg; Bera, François ULg; Thonart, Philippe ULg

in Journal of the Institute of Brewing (2012), 118(3), 295-305

Previous work having shown that Bacillus-subtilis-S499-based biocontrol treatments applied without aeration at the steeping stage of red sorghum malting offered good mould reduction but yielded malts with ... [more ▼]

Previous work having shown that Bacillus-subtilis-S499-based biocontrol treatments applied without aeration at the steeping stage of red sorghum malting offered good mould reduction but yielded malts with low levels of key hydrolytic enzymes, we attempted to raise these levels by aerating the steeping liquor, varying the steeping time (from 8 to 40 h) and temperature (from 25 to 35°C), and combining a biocontrol treatment with prior steeping in 0.2% NaOH. Aeration proved particularly important whenever B. subtilis cells were present in the steep liquor. The optimal temperatures for α- and β-amylase were 30 and 25°C, respectively. By increasing the steeping time it was possible to improve the α-amylase activity, but the -amylase activity peaked sharply between 16 and 20 h, depending on the steeping medium. A good compromise was steeping in biocontrol medium for 14 to 16 h at 30°C. Combination steeping treatments (0.2% NaOH for 8 h followed by biocontrol for 8 h) yielded malts of a quality approaching that afforded by dilute alkaline treatment. [less ▲]

Detailed reference viewed: 30 (8 ULg)
Full Text
Peer Reviewed
See detailOptimizing scale search in species distribution models
Serckx, Adeline ULg; Huynen, Marie-Claude ULg; Beudels-Jamar, Roseline et al

in Folia Primatologica : International Journal of Primatology = Internationale Zeitschrift für Primatologie = Journal international de Primatologie (2013), 84

The influence of spatial scale on ecological processes and pattern formation, such as species distribution is a major research topic since decades. It has become even more relevant in the context of ... [more ▼]

The influence of spatial scale on ecological processes and pattern formation, such as species distribution is a major research topic since decades. It has become even more relevant in the context of global change. In many studies the influence of a predictor on a response derived over multiple and discrete spatial scales is evaluated. Due to inherent issue of multiple testing, this approach can be problematic. In this study on bonobos distribution, we suggest a 3-steps procedure that overcomes this problem. It takes into account the decay of a predictor by using a weighting function of distance to the observation of interest. We use variables to account for human pressure, food availability, patchy structure of the forest and nesting sites re-use. In a first step, we run a model with a fixed scale for every predictor based on expert opinion. For significant variables, we derive then in a second step the distance weighted influence over a range of scales. This helps to narrow down the search for the final model parameter estimates. Findings indicate that bonobo distribution is driven on the intermediate scale by forest patches structure. Food availability explains their abundance only at smaller scales. Those results demonstrate the sadly well-known influence of habitat fragmentation on animals’ density and distribution but also highlight the importance to understand influences of scale, the animal perceptions of their environment, by using appropriate statistical procedures. Our method can be particularly useful to formulate specific management hypotheses for conservation. Furthermore, its principles can be of use to other types of studies, such as behavior. [less ▲]

Detailed reference viewed: 22 (5 ULg)
Full Text
Peer Reviewed
See detailOptimizing team members communication skills in radiation therapy : content and feasibility of a Belgian Interuniversity Curriculum
Caps, E.; Libert, Y.; Marchal, S. et al

in Psycho-oncology (2009, June), 18 (Suppl. 2)

Detailed reference viewed: 14 (0 ULg)
Peer Reviewed
See detailOptimizing the geometrical accuracy of 2D curvilinear finite element meshes
Remacle, Jean-François; Lambrechts, Jonathan; Toulorge, Thomas et al

Conference (2014, June 27)

Detailed reference viewed: 8 (0 ULg)