MASCOT 2014 Meeting


The MASCOT NUM 2014 meeting has been organized at ETH Zürich (Switzerland), from April 23rd to April 25th, 2014.

The best student presentation award has been conferred to Jonas Sukys (ETH Zürich) for his talk on Multi-level Monte Carlo finite volume methods for stochastic systems of hyperbolic conservation laws. He has won a prize of 1000€ to be used to go to a meeting. LOL

Program and handouts

Full conference brochure and abstracts.

Day 1 : April 23rd, 2014

08:30 Registration
09:15 Opening talk (Bruno Sudret, David Ginsbourger) Slides
09:30 Opening lecture (Claudia Schillings, ETH Zürich) Exploiting sparsity in Bayesian inverse problems of parametric operator equations Slides
10:30 Coffee break
10:50 S. Sukys (ETH Zürich) Multi-level Monte Carlo finite volume methods for stochastic systems of hyperbolic conservation laws Slides
11:20 D. Jacquemart (Université Rennes 1) Adaptive particle methods for rare event simulation in a Markovian framework
11:50 J. Nagel (ETH Zürich) Bayesian multilevel model calibration for inversion of “perfect” data in the presence of uncertainty Slides
12:20 Lunch break
13:30 Poster session
15:00 A. Ahidar (Université Paul Sabatier) Multivariate quantile surfaces and application to an aircraft problem
15:30 R. Fischer (Université Paris-Est) Modeling dependence under constraint
16:00 Coffee break
16:20 F. Zertuche (Université Joseph Fourier) Multi-fidelity regression using a non-parametric relationship
16:50 S. Nanty (Université Joseph Fourier) Uncertainty quantification and visualization for functional random variables Slides

Day 2: April 24th, 2014

09:00 Conference opening W. Kröger (ETH Risk Center)
09:15 B. Iooss & F. Mangeant MASCOT-NUM news Slides
09:30 F. Nobile (EPF Lausanne) Discrete least-square polynomial approximations for high dimensional uncertainty propagation Slides
10:30 Coffee break
10:50 J.P. Vert (Institut Curie) Machine Learning for personalised genomics Slides
11:50 A. Krause (ETH Zürich) Focusing exploration with confidence
12:50 Lunch break
14:15 F.S. Koutsourelakis (TUM) Simulation-based, high-dimensional stochastic optimization: Application in robust topology optimization under large material uncertainties Slides
15:15 S. Marelli (ETH Zürich) UQLab: A framework for uncertainty quantification in Matlab
16:00 Coffee break
16:20 D. Ginsbourger (University of Bern) Incorporating structural priors in Gaussian random field models Slides

Day 3: April 25th, 2014

09:00 P. Koumoutsakos (ETH Zürich) Data driven, molecular dynamics for nanoscale fluid mechanic
10:00 C. Chevalier (University of Bern), Y. Richet (IRSN), O. Roustant (EMSE) Presentation of the ReDICE consortium
10:30 Coffee break
10:50 P. Renard (Université de Neuchâtel) Multiple-point statistics as a tool to assess complex spatial uncertainty Slides
11:50 J. Irving (University of Lausanne) Stochastic inverse methods for near-surface geophysical problems Slides
12:50 Lunch break
14:00 J. Wiart (Orange Labs) Contribution of Statistics to the numerical assessment of the electromagnetic fields human exposure Slides
15:00 Closure

Poster session

G. Antinori (TU Munich). Robust design and optimization of a jet engine low pressure turbine rotor.

Y. Bennani (Université de Nice). Kriging for non-parametric ML estimation from region-censored data.

F. Franzelin (University of Stuttgart). Uncertainty quantification in Crack Propagation Simulations. Poster

J. Fruth (TU Dortmund). Sequential sensitivity analysis for computer experiments with functional inputs. Poster

S. Hamza (Université de Haute-Alsace). Sensitivity analysis of tire model microcoefficients. Poster

P. Kersaudy (Orange Labs). Use of chaos polynomials in a universal Kriging model: Application to the numerical dosimetry.

H. Maatouk (École des Mines de Saint Étienne). Gaussian processes for computer experiments with monotonicity information. Poster

V. Moutoussamy (Université Paul Sabatier). Estimation of rare events probability and quantiles under monotonicity constraints.

E. Musharbash, F. Nobile, T. Zhou (EPF Lausanne). Dynamical low rank approximation of time dependent PDEs with random data. Poster

R. Gantner (ETH Zürich). Multilevel Monte Carlo for Bayesian inverse problems.

R. Sala (Universita degli Studi di Firenze). An intuitive variance based variable screening method. Poster

R. Schöbi (ETH Zürich). PC-Kriging: Combining Polynomial Chaos Expansions and universal Kriging. Poster

S. Miro (Universität Bochum). System identification, meta-modeling, global sensitivity analysis and probabilistic analysis of failure in mechanized tunneling.

F. Tesei (EPF Lausanne). Multi level Monte Carlo methods with control variate for elliptic Stochastic Partial Differential Equations.

C. Walter (Université Paris VII) Getting into the failure domain: application in estimation of extreme quantiles and probabilities. Poster

O. Zahm (Ecole Centrale de Nantes). Goal-oriented low-rank approximations for high dimensional stochastic problems. Poster

Archives

More information is available on the ETH-hosted web site: http:www.mascotnum2014.ethz.ch

Call for PhD students&

The information about the call for abstracts for the PhD students' papers is available here.

Invited speakers

The list of invited speakers is available here.

Registration and Accomodation

The information about registration is available here.

The information about accommodation is available here.

Contacts

Organization: GDR MASCOT NUM Board, Bruno Sudret (ETH Zürich), David Ginsbourger (University of Bern).