Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Geological modeling --- Computer simulation

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In the field of shape optimization for CFD computations, users of geometry-modeling software normally have a wide variety of tools to choose from in order to parametrically design a hull from scratch and build a Fully Parametric Model (FPM). However, in practice most hull models are initially designed with no parametrization, and are then exported and imported between different software, normally forcing a subsequent required optimization to be performed in a Partially Parametric Model (PPM). These PPMs can sometimes be too complex for an average designer to build, since he may not be familiar with the mathematical constraints needed for the application of the required transformations, or the particular tools for given software to be used for the parametrization. In this context, a solution that can be introduced is sketched parametric modeling, which is the combination of complex geometry-modeling operations into intuitive, simple and user-friendly tools. By experience or input from adjoint/shape sensitivity analyses, designers may know where they would like their hull to be modified in order to improve CFD performance, and with the help of sketching, may parameterize those changes quickly to be input for a CFD optimization. This thesis presents an application of sketched parametric modeling in the parametric geometry-modeling and CFD integration software CAESES, and further use of it in a hull shape resistance optimization case using CFD viscous flow software FINE/Marine.

sketched parametric modeling, parametric modeling, CFD, hull shape optimization, ship resistance --- Ingénierie, informatique & technologie > Ingénierie civile

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Doing inference on a model defining an implicit likelihood that is not known in closed form is called likelihood-free inference. This occurs frequently in engineering and science domains where a simulator is used as a generative model of data, but the likelihood of the generated data is not known and is intractable. Given observed data, we combine the idea of hierarchical Bayesian modeling, empirical Bayes, and neural density estimation with normalizing flow to first learn a surrogate approximation of the model likelihood and then, to learn a prior distribution over the model parameters. The learned prior and the surrogate likelihood further allow to learn a posterior distribution for each observation. This is a general approach to likelihood-free inference, and is especially useful in settings where the simulator is too costly to run at inference time. We show the applicability of our methods on a real physical problem from high energy physics (HEP).

HEP, Amortized inference, Generative modeling --- Ingénierie, informatique & technologie > Ingénierie civile

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Dopamine is a chemical released by the brain which has long been associated with the pleasant feeling that accompanies a reward. The neurons that release such chemical, called dopaminergic neurons, have therefore been the subject of research and study for years. In fact, understanding the behaviour of the neurons that drive the regulation of dopamine levels and their interactions with the other entities that compose the brain is necessary for the understanding of a larger entity called the reward circuit. This circuitry which drives multiple phenomenona such as motivation, emotions, etc. Impairment and alterations in the circuitry is also known to lead to psychiatric disorders and addiction. 

Dopaminergic neurons present a specific behaviour as they actually fire in different modes which are imbricated together. On one hand, when unsolicited, they fire at a slow and robust rate. On the other hand, when they are triggered, their frequency of firing increases which also increases the dopamine release. This variability allows to regulate the dopamine in the brain.


This thesis focuses on the study of a model developed by G. Drion on the dynamics of dopaminergic activity. 

The first part of this thesis aims to reproduce experimental results that were obtained to validate the model. Using engineering tools such as model reduction and phase plane analysis, a deeper study of the dynamics of the model is performed in order to understand the mechanisms that drive the behaviour of the model. 

As a second part, the aim is to use the model in order to develop a hypothesis on the regulation of the firing frequency of dopaminergic neurons according to physiological properties of the neurons and linking it to the quantification of reward in the brain.

dopamine --- neuron --- modeling --- Ingénierie, informatique & technologie > Multidisciplinaire, généralités & autres

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

I. Study of data reduction and analysis of ACS/HST in UV line and continuum data for low-mass accreting T Tauri stars.II. Definition and implementation of a global software simulation package to assess and evaluate the performances of various detectors (CCD, CMOS,...) for high-precision measurements of various types of stars, applicable to both space and ground-based instrumentation. Application of this tool to the science detectors of the ESA M3 selected mission PLATO.

681.3*I63 <043> --- 520.2 --- Applications (Simulation and modeling)--Dissertaties --- Telescopes --- Theses --- Academic collection --- 520.2 Telescopes --- 681.3*I63 <043> Applications (Simulation and modeling)--Dissertaties