Topic selection & color key :

Topics

All topics are shown.

Forms of presentation

Forms of presentation

All types are shown.

SIAM Conference on Uncertainty Quantification (UQ20)

Wednesday – 25.03.2020

08:30

iCal
Zachary del Rosario | Stanford University | United States

Chun Yui Wong | University of Cambridge | United Kingdom

Laura White | NASA Langley Research Center | United States

Abhinav Gupta | Massachusetts Institute of Technology (MIT) | United States

Cécile Haberstich | CEA/DAM | France

Alec Dektor | UC Santa Cruz | United States

CT17: Dimension Reduction

Room:
MW HS 2001

Topic:
High-dimensional approximation

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Physics-informed Inference: Dimensional Analysis as Dimension Reduction

08:50

Embedded Ridge Approximations: Constructing Ridge Approximations Over Localized Scalar Fields For Improved Simulation-centric Dimension Reduction

09:10

- CANCELED - Uncertainty Quantification of High-dimensional Input and Output Spaces

09:30

Reduced-Dimension Bayesian Learning Machines for Discovering Dynamical Ocean Model Functions

09:50

Principal component analysis and boosted optimal weighted least-squares method for learning tree tensor networks

10:10

Hierarchical tensor methods for high-dimensional nonlinear PDEs

08:30

iCal
Md. Nurtaj Hossain | Indian Institute of Science, Bangalore | India

Felix Newberry | University of Colorado Boulder | United States

Christopher Müller | TU Darmstadt | Germany

Yanjin Wang | Institute of Applied Physics and Computational Mathematics, Beijing | China

Hannah Lu | Stanford University | United States

Daniel Meyer | ETH Zurich | Switzerland

CT20: ROM and Surrogate Models

Room:
MW HS 0001

Topic:
Reduced order models

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Uncertainty quantification of stochastic dynamical systems through efficient reduced order models using new error bounds

08:50

Parametric Low-fidelity Model Design for Bi-fidelity Approximation

09:10

A reduced basis method for elliptic PDEs with random data based on adaptive snapshots

09:30

- CANCELED - A Sampling Method for Kriging Surrogate Model with QOI Filling Space Based On Max-Min Distance

09:50

Predictive Accuracy of Dynamic Mode Decomposition

10:10

Surrogate Model Construction with Regression Element Trees

08:30

iCal
Vivak Patel | University of Wisconsin - Madison | United States

Arthur Macherey | Ecole Centrale de Nantes | France

Bruno Barracosa | EDF R&D | France

Mickaël Rivier | INRIA Saclay / CMAP, Ecole polytechnique | France

Dingcheng Luo | Oden Institute for Computational Engineering and Sciences, The University of Texas at Austin | United States

Ron Bates | Rolls-Royce plc | United Kingdom

CT11: Optimization

Room:
MW HS 1801

Topic:
Optimization and optimal control under uncertainty

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Connecting Kaczmarz to Kalman by Randomized Orthogonal Projections

08:50

Stochastic algorithm for the maximization of the expectation of a parameterized random variable

09:10

Bayesian methods for multi-objective simulation-based optimization: a review with some new ideas

09:30

- CANCELED - Constrained Multi-Objective Optimization under Uncertainty with Multi-fidelity Approximations

09:50

Taylor Approximations with Gaussian Mixtures for Optimization Under Uncertainty - A Multi Local Fidelity Approach

10:10

- CANCELED - Gaussian Processes and Sequential Design for Global Optimisation

08:30

iCal
Ilja Kröker | University of Stuttgart | Germany

Benjamin Phillips | NASA Langley Research Center | United States

Nora Lüthen | Chair of Risk, Safety and Uncertainty Quantification, ETH Zürich | Switzerland

Prem Ratan Mohan Ram | TU Braunschweig, Institute of Dynamics and Vibrations | Germany

Thomas West | NASA Langley Research Center | United States

CT03: Sparse grids and polynomial chaos

Room:
MW HS 0350

Topic:
High-dimensional approximation

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Arbitrary Multi-Resolution Multi-Wavelet-based Polynomial Chaos Expansion for Data-Driven Uncertainty Quantification

08:50

- CANCELED - Approaches for adaptive sampling with sparse grids for polynomial chaos expansion

09:10

Efficient computation of sparse polynomial chaos surrogates

09:50

Approximation of frequency response functions with the multi element generalized polynomial chaos method

10:30

- CANCELED - Uncertainty Quantification of Discontinuous Responses using Stochastic Expansions

08:30

iCal
João Reis | INRIA Saclay / CMAP, Ecole polytechnique | France

Ivana Pultarova | Czech Technical University in Prague, Faculty of Civil Engineering | Czech Republic

Alex Bespalov | University of Birmingham | United Kingdom

David Silvester | University of Manchester | United Kingdom

Shigetaka Kawai | The University of Tokyo | Japan

Pelin Çiloğlu | Middle East Technical University | Turkey

CT15: Stochastic Galerkin Methods and Iterative Solvers

Room:
MW HS 0250

Topic:
Uncertainty propagation

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Sample-Dependent Schwartz Preconditioners for Stochastic Elliptic Equations

08:50

New two-sided guaranteed spectral bounds for block preconditioning of stochastic Galerkin problems

09:10

A novel preconditioning technique for stochastic Galerkin finite element discretisations

09:30

Stochastic Galerkin mixed finite element approximation for Biot's consolidation model with uncertain inputs

09:50

Polynomial Annihilation-Based Stochastic Galerkin Method for Discontinuous System Responses in Aerodynamic Problems

10:10

Stochastic Discontinuous Galerkin Methods for Convection Diffusion Equations

08:30

iCal
David John | Corporate Research, Robert Bosch GmbH / Heidelberg University | Germany

Rebekah White | North Carolina State University | United States

Robert Barthorpe | The University of Sheffield | United Kingdom

Kathleen Pele | Ecole Centrale de Marseille | France

Mahesh D. Pandey | University of Waterloo | Canada

Angela Mihai | Cardiff University | United Kingdom

Amirreza Khodadadian | Leibniz University of Hannover | Germany

CT05: UQ in Material Science and Engineering Applications

Room:
MW HS 1250

Topic:
UQ for complex systems

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Hierarchical surrogate-based Approximate Bayesian Computation for an electric motor test bench

08:50

Quantifying Uncertainty in Warhead Design: How machining uncertainty affects volume and center of mass

09:10

Incorporating Uncertainty within a Forward Model Driven Structural Health Monitoring Paradigm

09:30

Development of a Markov chain model for the crack path prediction of heterogeneous materials

09:50

Quantifying parameter uncertainty in linear mixed-effects modeling of corrosion degradation process

10:10

- CANCELED - Likely instabilities in stochastic elasticity

10:30

- CANCELED - Bayesian Inversion for Variational Phase-Field Fracture Problems

08:30

iCal
Maarten Arnst | Université de Liège | Belgium

Ruanui Nicholson | University of Auckland | New Zealand

José Betancourt | Institut de Mathématiques de Toulouse - Ecole Nationale de l'Aviation Civile | France

Pierre Sochala | BRGM | France

Arunasalam Rahunanthan | Central State University | United States

Robert Lung | University of Edinburgh | United Kingdom

Aytekin Gel | Arizona State University | United States

CT08: UQ in Environmental Applications

Room:
MW 0608m

Topic:
UQ for complex systems

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

A mortar finite element method with embedded ensemble propagation for marine ice-sheet flow

08:50

Estimation of the basal sliding coefficient in ice sheet flow problems with uncertain rheological parameters

09:10

Kriging metamodeling of functional-inputs computer code for coastal flooding hazard assessment

09:30

A polynomial chaos framework for probabilistic predictions of storm surge events

09:50

- CANCELED - Uncertainty quantification of subsurface properties and the forecasting of aquifer contamination

10:10

Uncertainty Quantification for the inverse atmospheric dispersion problem with optical measurements

10:30

- NEW - The optimization and UQ tool box Nodeworks

08:30

iCal
Hiromichi Nagao | The University of Tokyo | Japan

Michael Liem | Institute of Fluid Dynamics, ETH Zurich | Switzerland

Pasha Piroozmand | ETH Zurich | Switzerland

Yuto Miyatake | Osaka University | Japan

Mustafa Mohamad | Courant Institute for Mathematical Sciences | United States

Ying Liang | The Chinese University of Hong Kong | Hong Kong

Raul Astudillo | Cornell University | United States

CT04: Data assimilation, inverse problems

Room:
MW ZS 1050

Topic:
Data assimilation

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Uncertainty quantification in data assimilation based on the four-dimensional variational method

08:50

Two-Stage Data Assimilation of Isolated Large Fractures in Reservoir Simulation Based on Ensemble Kalman Filters

09:10

Discrete adjoint based Data Assimilation for RANS Closure Models: The effect of the coupling between adjoint variables

09:30

Adjoint-based computation of the exact Hessian-vector multiplication

09:50

Predicting the Eulerian kinetic energy spectrum from Lagrangian drifters using combined data assimilation and parameter estimation

10:10

Improving well-posedness and robustness to data noisiness for electrical impedance tomography inverse problem

10:30

Solving Inverse Problems Using Bayesian Optimization of Composite Functions

08:30

iCal
Peter Marcy | Los Alamos National Laboratory | United States

Paul Gardner | University of Sheffield | United Kingdom

Kerry Klemmer | Princeton University | United States

Michael Schick | Corporate Research, Robert Bosch GmbH | Germany

Ruili Wang | Institute of Applied Physics and Computational Mathematics, Beijing | China

Tran Vi-vi Élodie Perrin | Ecole des Mines de Saint Etienne | France

Matieyendou Lamboni | Université de Guyane | French Guiana

CT18: Model Error and Sensitivity Analysis

Room:
MW ZS 2050

Topic:
Model error, discrepancy and calibration

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

- CANCELED - On Calibration of Parameter-Only Computer Models

08:50

Accounting for Model Discrepancy: A Sequential Bayesian History Matching and Gaussian Process Approach

09:10

Implied Models Approach for Turbulence Model Form Physics-Based Uncertainty Quantification

09:30

Physics informed neural networks for estimating simulation model credibility

09:50

Model Calibration: A physics-informed, Time-dependent Surrogate Method based on cylinder test

10:10

Global sensitivity analysis for numerical model with high-dimensional spatial output including strong discontinuities

10:30

Derivative-based decomposition of functions: dimension-free computation of total indices

08:30

iCal
Rodolfo de Freitas | Federal University of Rio de Janeiro | Brazil

Hans Yu | University of Cambridge | United Kingdom

Himaghna Bhattacharjee | University of Delaware | United States

Erika Hausenblas | Montanuniversität Leoben | Austria

Karen Larson | Brown University | United States

Ali Daher | Massachusetts Institute of Technology | United States

Robert Epp | Institute of Fluid Dynamics, ETH Zurich | Switzerland

CT06: UQ in Chemical Engineering and Biology

Room:
MW ZS 1450

Topic:
UQ for complex systems

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

An Embedded Deep Learning Model Discrepancy for Uncertainty Quantification on Combustion Computational Simulation

08:50

Time-Accurate Calibration of a Thermoacoustic Model on Experimental Images of a Forced Premixed Flame

09:10

- CANCELED - Quantifying Uncertainty In Multi-Scale Catalysis Models

09:30

Stochastic Systems in Biochemical Systems

09:50

Origin identification and uncertainty quantification for epidemic spread on networks

10:10

Bayesian Learning Machines for Glioblastoma Multiforme Brain Tumor Evolution

10:30

Adjoint-based parameter estimator for highly unsteady blood flow in the brain vasculature

08:30

iCal
Christiane Adcock | Stanford University | United States

Jakob Duerrwaechter | Universität Stuttgart | Germany

James Warner | NASA Langley Research Center | United States

Dongwei Ye | University of Amsterdam | Netherlands

Xueyu Zhu | University of Iowa | United States

Thomas Prescott | University of Oxford | United Kingdom

Philippe Blondeel | KU Leuven | Belgium

CT12: Multilevel and Multifidelity Monte Carlo

Room:
MW ZS 1550

Topic:
Statistical methods for UQ

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Multilevel Monte Carlo Sampling on Heterogeneous Computer Architectures

08:50

Efficient Automatized Iterative Multilevel and Multifidelity Monte Carlo Simulations for the Compressible Navier Stokes Equations

09:10

Generalized Multi-Model Monte Carlo Simulation for Uncertainty Propagation

09:30

Multi-Fidelity Monte Carlo with Semi-intrusive Algorithm for Multi-scale Simulation

09:50

Multifidelity Model-Informed Neural Network in Reduced Order Modeling

10:10

Multifidelity Approximate Bayesian Computation

10:30

h- and p- Multilevel Monte Carlo Methods in Geotechnical Engineering

08:30

iCal
Jonas Jehle | BMW Group | Germany

Biswarup Bhattacharyya | Univ Lyon, Université Claude Bernard Lyon 1, IFSTTAR, LBMC UMR_T9406 | France

Riccardo Tosi | Centro Internacional de Métodos Numéricos en Ingeniería (CIMNE) | Spain

Adrien Hirvoas | IFP Energies Nouvelles | France

Akiyoshi Yoshimura | Tokyo University of Technology | Japan

Camilo F. Silva | Technical University of Munich | Germany

Niklas Georg | Institut für Dynamik und Schwingungen, TU Braunschweig and Centre for Computational Engineering, TU Darmstadt | Germany

CT07: UQ in Transport and Energy Engineering Applications

Room:
MW 2250

Topic:
UQ for complex systems

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Creating a framework for uncertainty quantification in automotive crash simulations

08:50

- CANCELED - A surrogate approach for stochastic modeling of a crash box under impact loading in the time domain

09:10

Scalable dynamic asynchronous Monte Carlo framework applied to wind engineering problems

09:30

Recursive Bayesian Filtering procedures for parameter estimation of a wind turbine numerical model

09:50

- CANCELED - Restoration of Three-Dimensional Railway Track Geometry by a Recursive Nonlinear Filtering based on the Bayesian approach

10:10

Efficient uncertainty quantification of the acoustic radiation produced by HVAC devices via intrusive generalized chaos expansion

10:30

Enhanced Generalized Polynomial Chaos Approximation for Nanoplasmonics based on Conformal Maps

08:30

iCal
Yong Zeng | University of Missouri - Kansas City and National Science Foundation (NSF) | United States

Hannes Vandecasteele | KU Leuven | Germany

Roland Pulch | Universität Greifswald | Germany

Changqing Cheng | Binghamton University | United States

Shuai Guo | Technical University of Munich | Germany

Daniz Teymouri | Hong Kong University of Science and Technology | Hong Kong

CT14: Stochastic Dynamics and Multiscale Analysis

Room:
MW 1701

Topic:
Probability Theory for UQ

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

- CANCELED - Real-time Stochastic Volatility Tracking via Filtering for a Partially-Observed Heston Model

09:10

A micro-macro Markov chain Monte Carlo method for molecular dynamics using reaction coordinate proposals

09:30

System Norms for Sensitivity Analysis of Random Differential-Algebraic Equations

09:50

- CANCELED - Sequential Optimal Design for Complex Systems with Time Delay Under Uncertainty

10:10

Robust Identification of Frequency Response Function via a Multi-Fidelity Approach

10:30

- CANCELED - Uncertainty Quantification and Propagation of Partially Unknown Dynamical Systems using Monitoring Data

08:30

iCal
Danny Williamson | University of Exeter | United Kingdom

Anu Kauppi | Finnish Meteorological Institute | Finland

Anna Nikishova | University of Amsterdam, Informatics Institute | Netherlands

Omid Sedehi | Hong Kong University of Science and Technology | Hong Kong

Jonny Proppe | Georg-August University of Göttingen | Germany

Giulio Del Corso | Gran Sasso Science Institute | Italy

CT19: Modeling and Meta-modeling

Room:
MW HS 2235

Topic:
Model error, discrepancy and calibration

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Process based climate model development harnessing uncertainty quantification

08:50

Uncertainty Quantification in remote sensing: Model selection and imperfect forward modelling

09:10

Comparison of the non-intrusive and semi-intrusive metamodeling in uncertainty quantification

09:30

- CANCELED - Practical Insights on Hierarchical Bayesian Uncertainty Quantification and Propagation

09:50

Uncertainty-guided model design for transferable predictions of molecular properties

10:10

On the effect of the Purkinje network in the human electrophysiology: model validation and UQ analysis

08:30

iCal
Giovanni Rabitti | Department of Decision Sciences, Bocconi University, Milan | Italy

Sergey Oladyshkin | University of Stuttgart | Germany

Aikaterini Kyprioti | University of Notre Dame | United States

Daniela Jaruskova | Czech Technical University | Czech Republic

Panagiotis Demis | University of Surrey | United Kingdom

Jörg Buchwald | Helmholtz Zentrum für Umweltforschung GmbH - UFZ | Germany

CT09: Design of Experiments

Room:
MW HS 0337

Topic:
Design of experiments

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Simulator Inputs Screening with One-Factor-at-a-Time Designs

08:50

The connection between Bayesian Inference and Information Theory for model selection, information gain and experimental design

09:10

Adaptive Design of Experiments for Global Surrogate Modeling through Cross-Validation information

09:30

Effect of random parameters in nonlinear regression on an optimal experimental design

09:50

Bioprocess design space identification using constrained global sensitivity analysis

10:10

Application of experimental design-based assisted history matching for uncertainty quantification in radioactive waste repositories

08:30

iCal
Matej Leps | Czech Technical University in Prague | Czech Republic

Nadine Berner | Gesellschaft für Anlagen- und Reaktorsicherheit (GRS) gGmbH | Germany

Gabriel Sarazin | ONERA - The French Aerospace Lab | France

Patrick Héas | INRIA | France

Fritz Harland Sihombing | Ulsan National Institute of Science and Technology | Korea, Republic of

Dimitrios Patsialis | University of Notre Dame | United States

CT10: Rare Events and Risk

Room:
IAS 0.001

Topic:
Rare events and Risk

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Integrating Subset Simulation with Asymptotic Sampling for Estimating Small Failure Probabilities

08:50

An agent-based approach to explore complex dynamics of engineered systems for probabilistic safety analyses

09:10

Reliability-oriented sensitivity analysis in presence of data-driven epistemic uncertainty

09:30

- CANCELED - Selecting Reduced Models in the Cross-Entropy Method

09:50

- CANCELED - Probabilistic earthquake risk assessment using deep learning

10:10

Integration of reduced order modelling and multifidelity Monte Carlo simulation for efficient seismic risk assessment

08:30

iCal
Jonas Nitzler | Technical University of Munich | Germany

Torsten Enßlin | Max Planck Institute for Astrophysics | Germany

Victor Churchill | Dartmouth College | United States

Duc-Lam Duong | University of Sussex | United Kingdom

CT01: Bayesian inversion: theory

Room:
IAS 4.001

Topic:
Inverse problems

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

A generalized Bayesian multi-fidelity framework for high-dimensional inverse problems and uncertainty quantification

08:50

Information field theory: recovering fields & their uncertainty from data and knowledge

09:30

Total variation Bayesian learning via synthesis

10:10

Bayesian inverse problems in scalar conservation laws

08:30

iCal
Nishan Mudalige | University of Guelph | Canada

Louise Kimpton | University of Exeter | United Kingdom

Nevin Martin | Sandia National Laboratories | United States

Anouar Meynaoui | INSA de Toulouse | France

Kabir Olorede | Kwara State University | Nigeria

Yoonsang Lee | Dartmouth College | United States

CT21: Statistical Methods for UQ

Room:
IAS 4.002

Topic:
Statistical methods for UQ

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

[ Moved from MW HS 0337 ]

08:30

- CANCELED - A goodness of fit for exponential models over S^(p-1)

08:50

Correlated Bernoulli Processes Using De Bruijn Graphs

09:10

Estimation of physical tolerance bounds for functional data

09:30

Practical implementation of aggregated tests of independence based on dependence measures

10:10

A New Covariance Estimator for Sufficient Dimension Reduction in High-dimensional Data with Undersized Sample Problems

10:30

Non-parametric importance sampling for parameter estimation through MCMC

08:30

iCal
Agnimitra Dasgupta | University of Southern California | United States

Steven Atkinson | GE Research | United States

Olga Fuks | Stanford University | United States

Rishith Ellath Meethal | Siemens AG, Corporate Technology | Germany

Subhayan De | University of Colorado Boulder | United States

Sergei Kucherenko | Imperial College London | United Kingdom

CT16: Neural Networks and Machine Learning

Room:
Interims Lecture Hall 101

Topic:
Physics-informed and data-driven UQ methods

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:50

Physics-informed machine learning using inferred-physics models: application to developing models for a MR damper

09:10

Active learning in structure-free discovery of governing laws from data

09:30

Physics Informed Machine Learning for Nonlinear Transport in Porous Media

09:50

A FEM-informed neural network framework for uncertainty quantification

10:10

Neural Network Training using Bi-fidelity Data for Uncertainty Quantification

10:30

Application of machine learning and global sensitivity analysis for identification and visualization of design space

08:30

iCal
Hossein Mohammadi | University of Exeter | United Kingdom

Wei-Ann Lin | National Cheng Kung University | Taiwan

Ray-Bing Chen | National Cheng Kung University | Taiwan

Thierry Gonon | Ecole centrale de Lyon | France

Wei Chen | Northwestern University | United States

Baptiste Kerleguer | CEA DAM | France

CT13: Gaussian Processes

Room:
Interims Lecture Hall 102

Topic:
Surrogate models

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

Emulating computer models with step-discontinuous outputs using Gaussian processes

08:50

Tree-based Gaussian Process with Many Qualitative Factors

09:10

Bayesian Variable Selection in Gaussian Process Models for Computer Experiments

09:30

Sequential input adding in computer experiments

09:50

Alternative Latent Space Representations in Latent Variable Gaussian Process Modeling

10:10

Multifidelity Gaussian process metamodel for time-dependent outputs

08:30

iCal
Jonas Sukys | Eawag: Swiss Federal Institute of Aquatic Science and Technology | Switzerland

Daniel Waelchli | ETH Zurich | Switzerland

Yue Qiu | ShanghaiTech University | China

Anabel del Val | von Karman Institute for Fluid Dynamics | Belgium

Jiahua Jiang | Virginia Tech | United States

Samah El Mohtar | King Abdullah University of Science and Technology | Saudi Arabia

CT02: Bayesian inversion: applications and software

Room:
Exzellenzzentrum 0003

Topic:
Inverse problems

Form of presentation:
Contributed Lecture

Duration:
140 Minutes

08:30

SPUX - a Scalable Package for Bayesian Uncertainty Quantification

08:50

Korali: a high-performance framework for Bayesian uncertainty quantification and optimization

09:10

Randomized Low-rank Ensemble Kalman Filter for Nonlinear Networks

09:30

Robust Bayesian Inference under Limited Information and its Application to Atomic Spectra for Atmospheric Entry Systems

09:50

- CANCELED - Hybrid Projection Methods for Large-scale Inverse Problems with Mixed Gaussian Priors

10:10

- CANCELED - Bayesian Inference and Markov Chain Monte Carlo Sampling for Lagrangian Particle Tracking in the Ocean

11:15

iCal
Björn Sprungk | TU Bergakademie Freiberg | Germany

SIAG/UQ Early Career Prize

Chair(s)
Omar Ghattas (The University of Texas at Austin)

Fabio Nobile (Ecole Polytechnique Fédérale de Lausanne (EPFL))

Fabio Nobile (Ecole Polytechnique Fédérale de Lausanne (EPFL))

Room:
MW HS 2001

Topic:
Inverse problems

Form of presentation:
Plenary Lecture

Duration:
30 Minutes

11:15

Noise-Level Robust Sampling Methods for Bayesian Inverse Problems

11:15

iCal

SIAG/UQ Early Career Prize - streamed from HS 2001

Room:
MW HS 0001

Topic:
Inverse problems

Form of presentation:
Plenary Lecture

Duration:
30 Minutes

11:45

iCal
David Higdon | Virginia Polytechnic Institute and State University | United States

IP03: David M. Higdon: UQ and Bayesian Model Calibration Applied to Stochastic Systems

Room:
MW HS 2001

Topic:
Model error, discrepancy and calibration

Form of presentation:
Plenary Lecture

Duration:
45 Minutes

11:45

UQ and Bayesian Model Calibration Applied to Stochastic Systems

11:45

iCal

IP03 - streamed from HS 2001: David M. Higdon: UQ and Bayesian Model Calibration Applied to Stochastic Systems

Room:
MW HS 0001

Topic:
Model error, discrepancy and calibration

Form of presentation:
Plenary Lecture

Duration:
45 Minutes

12:30

iCal
**
Room :
**
IAS 4.001, IAS 4.002

JUQ Editorial Board Meeting - Upon invitation only!

**
Duration :
**
90 Minutes

14:00

iCal
Ben Adcock | Simon Fraser University | Canada

MT03: Ben Adcock: Practical approximation in high dimensions: from polynomials to neural networks

Room:
MW HS 2001

Topic:
High-dimensional approximation

Form of presentation:
Mini-tutorial

Duration:
120 Minutes

14:00

Practical approximation in high dimensions: from polynomials to neural networks

14:00

iCal
Michael Mascagni | Florida State University | United States

Abani Patra | Tufts University | United States

Gregory Kiar | Concordia University | Canada

Peter Coveney | University College London | United Kingdom

MS571: Reproducibility and Uncertainty Quantification

Room:
MW HS 0001

Topic:
Verification and validation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Reproducibility has emerged as an issue in experimental work, and as a consequence, computational work is now being scrutinized for reproducibility. Predating emphasis on reproducibility, computational results have developed methodologies for verification and validation, and more recently with techniques based on probabilistic analysis that are grouped together as Uncertainty Quantification. This minisymposium will be a venue for these ideas and techniques applied to computation.

14:00

The "White Rat" of Numerical Reproducibility

14:30

Reproducibility and UQ for Two Classes of Geophysical Models

15:00

Uncertainty quantification for neuroimaging data analyses

15:30

Reproducibility, computability, and the scientific method

14:00

iCal
Mathias Staudigl | Maastricht University | Netherlands

Andreas Van Barel | KU Leuven | Belgium

Caroline Geiersbach | University of Vienna | Austria

Philipp Guth | University of Mannheim | Germany

MS803: Optimization and Estimation of Complex Systems under Uncertainty Part III of III: Approximation

Chair(s)
Thomas M. Surowiec (Philipps-Universität Marburg)

Drew P. Kouri (Sandia National Laboratories)

Drew P. Kouri (Sandia National Laboratories)

Room:
MW HS 1801

Topic:
Optimization and optimal control under uncertainty

Form of presentation:
Mini-symposium

Duration:
120 Minutes

The overwhelming majority of modern applications in the natural sciences, engineering, and beyond require both statistical estimation to accurately quantify the behavior of unknown distributed parameters in complex systems as well as a means of making optimal decisions that are resilient to this uncertainty. In this minisymposium, we aim to connect researchers working in optimization of complex systems under uncertainty such as equilibrium problems, differential algebraic equations, and partial differential equations, with statisticians working in variational statistics, infinite-dimensional statistical estimation, and optimum experimental design.

14:00

Variance Reduction schemes for stochastic Variational inequalities

14:30

Performance of Multilevel Monte Carlo techniques for robust PDE constrained optimization

15:00

Stochastic Proximal Gradient Method in Hilbert Spaces

15:30

A Quasi-Monte Carlo Method for PDE-Constrained Optimization under Uncertainty

14:00

iCal
Bastian Bohn | University of Bonn | Germany

Ian Sloan | University of New South Wales | Australia

MS031: Kernel, Quasi-Monte Carlo, and Sparse Grid Methods for High-dimensional Approximation and Integration (Part I of III)

Chair(s)
Yoshihito Kazashi (EPFL)

Alexander Gilbert (Universität Heidelberg)

Fabio Nobile (EPFL)

Michael Griebel (University of Bonn and Fraunhofer SCAI)

Alexander Gilbert (Universität Heidelberg)

Fabio Nobile (EPFL)

Michael Griebel (University of Bonn and Fraunhofer SCAI)

Room:
MW HS 0350

Topic:
High-dimensional approximation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

This mini-symposium aims at bringing together people working on kernel and other sampling-based approximation methods for high-dimensional problems, in particular, but not restricted to, quasi-Monte Carlo methods, and sparse grid methods. Kernel methods and the related Gaussian Process surrogate models are a powerful class of numerical methods, and they are often employed in problems arising in uncertainty quantification. Nonetheless, there is much to be explored in their theoretical analysis for UQ applications, which are often formulated as high-dimensional approximation or integration problems.

On the other hand, the theory and applicability of QMC and sparse grid approximation/integration techniques in high or infinite dimensional problems have seen considerable advances in the last years, yet are far from addressing all problems of interest in UQ.

The objective of this mini-symposium is to showcase the latest theoretical results and exchange ideas on sampling-based high-dimensional integration and approximation methods targeting UQ applications.

14:00

Accelerating least-squares by chained approximations

14:30

- CANCELED - Kernel-based lattice point interpolation for uncertainty quantification using periodic random variables

14:00

iCal
Björn Sprungk | TU Bergakademie Freiberg | Germany

Aretha Teckentrup | University of Edinburgh | United Kingdom

Matteo Giordano | University of Cambridge | United Kingdom

Masoumeh Dashti | University of Sussex | United Kingdom

MS431: Would Hadamard have used Bayes' rule? - On robustness and brittleness of statistical inversion

Room:
MW HS 0250

Topic:
Inverse problems

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

In many scientific disciplines, researchers encounter inverse problems where observational data shall be used to calibrate mathematical models. Hadamard considered the solvability of such problems in terms of their "well-posedness". He called a problem well-posed, if a solution exists, if the solution is unique, and if the solution depends continuously on the data. Inverse problems are typically not well-posed (i.e., ill-posed) and require some regularization. Today's availability of high-performance computing has raised the popularity of statistical approaches to inverse problems and probabilistic regularizations; like the Bayesian approach.

In this minisymposium, we consider the robustness and non-robustness (that is, the brittleness) of Bayesian inverse problems and related approaches. This includes the robustness with respect to perturbations in the data (that is, the well-posedness), but also with respect to perturbations in the prior measure or the likelihood. Perturbations in the

prior also include a potential ill-specification of the prior model whereas perturbations in the likelihood include the replacement of the mathematical model by a discretised version or a surrogate. Moreover, we are interested in the robustness of algorithms used for statistical inversion, such as MCMC, particle filters, variational Bayes, and approximate Bayesian computation.

14:00

On the Local Lipschitz Robustness of Bayesian Inverse Problems

14:30

Gaussian process emulators in Bayesian inverse problems

15:00

- NEW - Consistency of Bayesian inference with Gaussian priors in an elliptic nonlinear inverse problem

15:30

MAP estimators and posterior consistency for Bayesian inverse problems for functions

14:00

iCal
Tim Reid | North Carolina State University | United States

Florian Schäfer | California Institute of Technology | United States

Takeru Matsuda | Department of Mathematical Informatics, The University of Tokyo | Japan

Alejandro Diaz | University College London | United Kingdom

MS021: Probabilistic Numerical Methods for Differential Equations and Linear Algebra (Part I of II)

Chair(s)
Philipp Hennig (University of Tübingen & Max Planck Institute for Intelligent Systems)

Alejandro Diaz (University College London)

Alejandro Diaz (University College London)

Room:
MW HS 1250

Topic:
Probability Theory for UQ

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In many important inverse problems and engineering computations -e.g. numerical weather prediction, medical tomography, reliability analysis- data are related to parameters of interest through the solution of an ordinary or partial differential equation (DE). To proceed with computation, the DE must be discretised and solved through linear algebra methods. However, such discretisation introduces bias into parameter estimates and can in turn cause conclusions to be over-confident. Probabilistic numerical methods for DEs and linear algebra aim to provide uncertainty quantification in the solution space of the DE to properly account for the fact that the governing equations have been altered through discretisation. In contrast to the worst-case error bounds of classical numerical analysis, the stochasticity in DEs and linear solvers serves as the carrier of uncertainty about discretisation error and its impact. This statistical notion of discretisation uncertainty can then be more easily propagated to later inferences, e.g. in a Bayesian inverse problem. Several such probabilistic numerical methods have been developed in recent years, and the connections and distinctions between these methods are starting to be modelled and understood. In particular, an important challenge is to ensure that such uncertainty estimates are well-calibrated. This minisymposium will examine recent advances in both the development and implementation of probabilistic numerical methods in general.

14:00

Prior Distributions and Test Statistics for the Bayesian Conjugate Gradient Method

14:30

A probabilistic view on sparse Cholesky factorization

15:00

Estimation of ordinary differential equation models with discretization error quantification

15:30

Probabilistic rare-event simulation

14:00

iCal
Hanno Gottschalk | University of Wuppertal | Germany

Andreas Stein | University of Stuttgart | Germany

Robin Merkle | University of Stuttgart | Germany

MS641: Random PDEs with Lévy Fields

Room:
MW 0608m

Topic:
Uncertainty propagation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

Uncertainty quantification plays an increasingly important role in a wide range of problems in the physical sciences and financial markets. The underlying model may be subject to various uncertainties such as parameter or domain uncertainty, model uncertainty, numerical errors, or some intrinsic stochastic variability of the model. In the latter case, the uncertainty could be either introduced by measuring instruments or is the result of insufficient observations. For realistic simulations in the underlying differential equation model this is reflected via a random operator and/or random data. These parameters are often modelled as space-time Gaussian processes, leading to continuous random functions and thin-tailed, symmetric normal distributions.

Although Gaussian random objects have convenient analytical properties, for several applications, however, it might be favourable to model the stochastic quantities as discontinuous fields or processes which also allow for asymmetric and heavy-tailed distributions. In this minisymposium we bring together researchers whose foci are on stochastic or random partial differential equations which are influenced by discontinuous fields or processes.

14:30

Elliptic PDE with Lévy coefficients: Integrability and Convergence of Approximations of Karhunen-Loève Type

15:00

A Fully Discrete Approximation Scheme for a Stochastic Transport Problem with Lévy Noise

15:30

Subordinated Gaussian Random Fields in Elliptic SPDEs

14:00

iCal
Giovanni M. Porta | Politecnico di Milano | Italy

Ana Gonzalez-Nicolas | University of Stuttgart | Germany

Fadji Z. Maina | Lawrence Berkeley National Laboratory | United States

Kyle R. Steffen | University of Texas at Austin | United States

MS061: Uncertainty Quantification in Hydrology (Part I of II)

Chair(s)
Gabriele Chiogna (Technical University of Munich and University of Innsbruck)

Ana Gonzalez-Nicolas (University of Stuttgart)

Ana Gonzalez-Nicolas (University of Stuttgart)

Room:
MW ZS 1050

Topic:
Uncertainty propagation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

Hydrological model simulations are often complicated by inevitable uncertainties in initial conditions, boundary conditions, and parameter fields. A proper identification and quantification of such uncertainties are nowadays a must for any modern hydrologist. In this mini symposium, beside presentations focusing on how uncertainty quantification can be properly performed for problems typical of hydrological sciences (e.g., flow and transport in porous media, river and karst spring discharge predictions, surface water-groundwater interaction…), we want to emphasize why uncertainty quantification is relevant in hydrology and its implication for engineering applications.

The minisimposium received funding from the International Graduate School of Science and Engineering of the Technical University of Munich.

14:00

Global sensitivity indicators for soil and groundwater contamination risk assessment

14:30

Optimal monitoring strategies for minimal uncertainties about a groundwater divide

15:00

Feedbacks between evapotranspiration and subsurface flow: a global sensitivity analysis

15:30

Inference and prediction of air-ice-ocean momentum transfer

14:00

iCal
Peer-Timo Bremer | Lawrence Livermore National Labs | United States

Kellin Rumsey | University of New Mexico | United States

Daniel O’Malley | Los Alamos National Laboratory | United States

Nishant Panda | Los Alamos National Laboratory | United States

MS471: Leveraging the Interplay Between UQ and ML for Mutual Benefit (Part I of II)

Chair(s)
Justin Newcomer (Sandia National Laboratories)

Gowri Srinivasan (Los Alamos National Laboratory)

Jean-Christophe Weill (CEA)

Gowri Srinivasan (Los Alamos National Laboratory)

Jean-Christophe Weill (CEA)

Room:
MW ZS 2050

Topic:
Physics-informed and data-driven UQ methods

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Machine Learning (ML) has evolved into a core technology in many scientific applications. Solutions often require large labeled datasets to achieve high model accuracy. Unfortunately, this is a major bottleneck for many scientific computing applications, where numerical simulations are very expensive. Training on limited data can lead to significant uncertainties or errors when invoked outside the training space. But the fast execution of ML models once trained also make them ideal for exploring large numbers of runs for Uncertainty Quantification (UQ). Furthermore, many popular ML methods lack the needed mathematical support to prove robustness and reliability to motivate their use in scientific computing and uncertainty quantification UQ applications. This two-part mini-symposium will explore the interplay between ML and UQ, focusing in the following areas: (1) How do we leverage ML successes for scientific computing problems with uncertain inputs? (2) How do we use UQ methods to assess ML predictions and augment them with uncertainty estimates, error bounds, or prediction intervals? Addressing challenges in these areas will lead to greatly improve predictive capabilities. Methods that incorporate mathematical and scientific principles for uncertainty estimates in ML are needed. Literature in statistics can be leveraged for improving the model validation process and advances in UQ and V&V will greatly enhance the mathematical and scientific computing foundations for ML.

14:00

Learning-by-Calibrating: Improved Surrogate Models using Calibration as a Training Objective

14:30

Embracing unidentifiability in Bayesian model calibration with modularization

15:00

Learning to regularize with a variational autoencoder for hydrologic inverse analysis

15:30

Data Driven Upscaling - Emulating Mesoscale Physics Using Machine Learning in Fractured Media

14:00

iCal
Eric Larour | Jet Propulsion Laboratory | United States

Andy Aschwanden | University of Alaska Fairbanks | United States

Kevin Bulthuis | Université de Liège | Belgium

Tamsin Edwards | King’s College London | United Kingdom

MS613: Uncertainty estimates of the cryosphere and its forcings (Part III of III)

Chair(s)
Nicole-Jeanne Schlegel (Jet Propulsion Laboratory, California Institute of Technology)

Andrew Davis (Cold Regions Research and Engineering Laboratory)

Andrew Davis (Cold Regions Research and Engineering Laboratory)

Room:
MW ZS 1450

Topic:
UQ for complex systems

Form of presentation:
Mini-symposium

Duration:
120 Minutes

The cryosphere and the processes that force its evolution have profound and permanent effects on the global climate. In particular, Arctic amplification leads to extreme mid-latitude weather and glacier and ice sheet retreat is increasing global mean sea level causing the ocean to encroach onto coastal communities. Despite potentially devastating impacts, accurate predictions of future dynamics and rigorous characterizations of the associated uncertainty remain elusive. Misunderstood physics and computational limitations require complex physical processes to be parameterized and calibrated using noisy data that is sparse in both space and time. However, collecting data in remote polar regions is difficult, dangerous, and expensive. Therefore, we must leverage remote sensing techniques and wisely allocate limited resources. Finally, predictive uncertainties must be quantified to give meaningful error bounds on quantities of interest, such as future mean sea level. This session discusses recent advancements trying to understand the dynamic processes governing the cryosphere given observations and/or models as well as techniques to obtain and analyze data.

14:00

Uncertainty quantification of sea-level projections informed by the Ice and Sea-level System Model

14:30

- CANCELED - Using Gaussian Process Emulators to reduce uncertainty in sea level projections with ice sheet models

15:00

Stochastic Modeling of Uncertainty in Ice Sheet Models, with application to the Antarctic Ice Sheet response to Climate Change

15:30

Quantifying the uncertainties of climate model predictions and ice sheet contributions to sea-level rise

14:00

iCal
Spencer Lunderman | University of Arizona | United States

Kayo Ide | University of Maryland, College Park | United States

Rishikesh Yadav | King Abdullah University of Science and Technology (KAUST) | Saudi Arabia

Marcus van Leir Walqui | NASA Goddard Institute for Space Studies | United States

MS791: Bayesian Inference in Earth Science (Part I of II)

Chair(s)
Spencer Lunderman (University of Arizona)

Derek Posselt (Jet Propulsion Laboratory, California Institute of Technology)

Matthias Morzfeld (University of Arizona)

Derek Posselt (Jet Propulsion Laboratory, California Institute of Technology)

Matthias Morzfeld (University of Arizona)

Room:
MW ZS 1550

Topic:
Data assimilation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

There is typically a mismatch between observations of a process, and its representation in a mathematical or numerical model. Such error arises because the model is incomplete or approximate, and errors are amplified by noise in the observations, as well as uncertain, or completely unknown, model states and parameters. In Earth science, errors of these types must be quantified, and a natural tool to do so is Bayesian inference, where errors are described via conditional probabilities defined for the model, its parameters, and the observations.

This mini-symposium will focus on the numerical solution of Bayesian inference problems in Earth sciences which are usually characterized by a large dimension (many parameters and states) and few observations (relative to the number of states and parameters). Moreover, Earth science applications require solutions to three types of Bayesian inference problems: state estimation (data assimilation), parameter estimation, and joint state and parameter estimation.

Our mini-symposium will showcase Bayesian inference "in action" in Earth science. It will provide an opportunity for interaction among applied mathematicians, interested in the numerics of Bayesian inference, and Earth scientists, who use Bayesian inference to break new ground in their respective fields.

14:00

- CANCELED - Simultaneous parameter and state estimation by derivative-free optimization of ensemble Kalman filter residuals

14:30

- CANCELED - Fitness of the ensemble approach in data assimilation system

15:00

- CANCELED - Bayesian hierarchical modeling of spatial rainfall extremes using rate mixtures

15:30

Challenges in Simulating Cloud and Precipitation Processes: Frontiers for Bayesian Inference, Model Selection, and Machine Learning

14:00

iCal
Peng Chen | Oden Institute, UT Austin | United States

Francois-Xavier Briol | University College London | United Kingdom

Uros Seljak | University of California at Berkeley | United States

MS111: Inference and preconditioning via Stein methods, flows, and other transport maps (Part I of II)

Chair(s)
Benjamin Peherstorfer (Courant Institute, New York University)

Youssef Marzouk (Massachusetts Institute of Technology)

Yaoliang Yu (University of Waterloo)

Youssef Marzouk (Massachusetts Institute of Technology)

Yaoliang Yu (University of Waterloo)

Room:
MW 2250

Topic:
Statistical methods for UQ

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

Transport maps are deterministic couplings between probability measures with broad applications in uncertainty quantification and machine learning. They have been used for posterior sampling in Bayesian inference, for accelerating Markov chain Monte Carlo and importance sampling algorithms, and as building blocks of generative models and density estimation methods. More broadly, transport---including but not limited to optimal transport---provides an important mathematical foundation for many tools in machine learning and uncertainty quantification. The recent surge of interest in transport maps has been accompanied by efficient numerical methods that make constructing and learning such maps tractable in high dimensions and for large data sets. This minisymposium brings together researchers from uncertainty quantification and machine learning to discuss recent advances in theory, numerics, and applications of transport maps and related techniques.

14:30

Projected Stein variational Newton: A fast and scalable Bayesian inference method in high dimensions

15:00

- CANCELED - Minimum Stein Discrepancy Estimators and their Gradient Flows

15:30

Generative models for Bayesian inference

14:00

iCal
Sophie Marque-Pucheu | Orange Gardens | France

Deyu Ming | University College London | United Kingdom

David Woods | University of Southampton | United Kingdom

Andrea Bevilacqua | Istituto Nazionale di Geofisica e Vulcanologia | Italy

MS281: Uncertainty quantification of multi-physics computer models

Room:
MW 1701

Topic:
UQ for complex systems

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Many modern simulators of physically realistic phenomena use multiple, heterogeneous sub-models, possibly involving different types of physical modelling and dimensionality. This usage poses many challenges that need to account for the links across sub-models: building surrogates, designing experiments, exploring sensitivities, reducing dimensions, etc. Both theoretical investigations and implementations are hampered by the complex nature of such models, and need to be tailored to the specific chain of models. In this mini-symposium, we present a series of talks that address such challenges and offer theoretical as well as practical solutions, together with illustrations. In particular, realistic models of geophysical and biological hazards often include feedbacks across sub-models or are combinations of sub-models of precursory phenomena – models that set the stage for dangerous events and can be informed by monitoring data – as well as the models of the hazardous phenomenon itself. These challenges require solutions that acknowledge the interactions across multi-physics components.

14:00

- CANCELED - An efficient dimension reduction for the Gaussian process emulation of two nested codes with functional outputs

14:30

Integrated Emulation of Systems of Simulators

15:00

Approaches to the Emulation of Chains of Computer Models with Application to Epidemic Policy Making

15:30

- CANCELED - Probabilistic hazard mapping in a volcanic field under rapidly evolving monitoring signals: integration of probability maps of vent opening location and physical models of pyroclastic density currents

14:00

iCal
Mohammad Motamed | The University of New Mexico | United States

Arthur Vidard | Laboratoire Jean Kuntzmann | France

Tobias Sutter | EPFL | Switzerland

Philippe Moireau | INRIA | France

MS251: Similarity measures and distances in forward and inverse UQ problems (Part I of II)

Chair(s)
Didier Lucor (LIMSI - CNRS)

Mohammad Motamed (The University of New Mexico)

Lionel Mathelin (LIMSI - CNRS)

Mohammad Motamed (The University of New Mexico)

Lionel Mathelin (LIMSI - CNRS)

Room:
MW HS 0337

Topic:
Other

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

[ Moved from MW HS 2235 ]

The analysis and comparison of dynamic objects and deforming shapes is important in many real-world applications. Examples include wildfire front-tracking problems, impulse propagation in cardiac tissues, tumor growth, oil reservoir and spill simulations, and pollutant plume dispersion, just to name a few. There are several difficulties that can make the analysis a daunting task and hence need to be addressed: 1) the problem is subjected to uncertainty in the location of structures due to numerical errors, measurement noise, and/or intrinsic variations in the system; 2) strong shape deformations and topological changes may not be well captured at all scales; and 3) the notion of distance or similarity between objects can be characterized in various ways.

This situation has fostered a recent body of work focused on both analytical and computational developments in metric spaces. As an example, the Wasserstein metric has become an increasingly popular tool in such diverse fields as image processing, optimization, neural networks, seismic imaging, and numerical conservation laws. It opens up promising avenues for uncertainty quantification, Bayesian inference and data assimilation, where robust comparisons and mappings between different probability measures are often needed.

This MS will review recent advances, applications and remaining challenges of tailored metric spaces and similarity measures for structure-sensitive uncertainty quantification and inference problems.

14:00

Wasserstein metric-driven Bayesian Inversion with applications to Wave Propagation Problems

14:30

Optimal transport-based for variational data assimilation

15:00

Optimal Decision-Making and Uncertainty Quantification — Beyond IID Data

15:30

Robust Kalman filtering of shape observations using shape metrics

14:00

iCal
Antoine Blanchard | Massachusetts Institute of Technology | United States

Oliver Schmidt | University of California, San Diego | United States

Pantelis Vlachas | ETH Zürich | Switzerland

MS101: Statistical Prediction and Quantification of Extreme Events in Complex Systems (Part I of II)

Chair(s)
Antoine Blanchard (Massachusetts Institute of Technology)

Themistoklis P. Sapsis (Massachusetts Institute of Technology)

Themistoklis P. Sapsis (Massachusetts Institute of Technology)

Room:
IAS 0.001

Topic:
Rare events and Risk

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Extreme events are short-lived episodes occurring due to exogenous causes or internal instabilities during which observables significantly depart from their mean values. A great deal of effort has been devoted to predicting and statistically quantifying extreme events because they can have catastrophic consequences (e.g., structural failure, rogue waves, extreme weather conditions, and market crashes). This is an arduous task because the systems that give rise to extreme events are most often highly complex and strongly nonlinear. This mini-symposium provides a venue to review the latest advances in the field.

14:00

Optimal Sampling with Gaussian Process Regression

14:30

A Conditional Space-Time POD Formalism for the Statistical Description of Intermittent and Rare Events

15:00

Recurrent Neural Networks and Reservoir Computing for Spatio-Temporal Forecasting of Chaotic Dynamics: A Comparative Study

14:00

iCal
Lorenzo Tamellini | Istituto di Matematica Applicata e Tecnologie Informatiche (CNR-IMATI) | Italy

Michael Eldred | Sandia National Laboratories | United States

Friedrich Menhorn | Technical University of Munich | Germany

Quentin Ayoul-Guilmard | EPFL | Switzerland

MS682: Multilevel and Multifidelity approaches for forward/inverse Uncertainty Quantification and optimization under uncertainty (Part II of III)

Chair(s)
Panagiotis Tsilifis (EPFL)

Gianluca Geraci (Sandia National Laboratories)

Alex Gorodetsky (University of Michigan)

John Jakeman (Sandia National Laboratories)

Juan Pablo Madrigal Cianci (EPFL)

Michael Eldred (Sandia National Laboratories)

Gianluca Geraci (Sandia National Laboratories)

Alex Gorodetsky (University of Michigan)

John Jakeman (Sandia National Laboratories)

Juan Pablo Madrigal Cianci (EPFL)

Michael Eldred (Sandia National Laboratories)

Room:
Interims Lecture Hall 101

Topic:
Uncertainty propagation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In the last decades, the advancements in both computer hardware/architectures

and algorithms enabled numerical simulations at unprecedented scales. In parallel,

Uncertainty Quantification (UQ) evolved as a crucial task to enable predictive

numerical simulations. Therefore, a great effort has been devoted in advancing the UQ algorithms

in order to enable UQ for expensive numerical simulations, however the combination of an extremely

large computational cost associated to the evaluation of a high-fidelity model and the presence of a moderate/large

set of uncertainty parameters (often correlated to the complexity of the numerical/physical assumptions)

still represents a formidable challenge for UQ.

Multilevel and multifidelity strategies have been introduced to circumvent these difficulties by

reducing the computational cost required to perform UQ with high-fidelity simulations. The

main idea is to optimally combine simulations of increasingly resolution levels or model fidelities

in order to control the overall accuracy of the surrogates/estimators. This task is accomplished by

combining large number of less accurate numerical simulations with only a limited number of high-fidelity,

numerically expensive, code realizations. In this minisymposium we present contributions related to the state-of-the-art in both forward and inverse multilevel/multifidelity UQ and related areas as optimization under uncertainty.

14:00

Recent Advances on IGA-based Multi-Index Stochastic Collocation

14:30

Recent advancements in multilevel-multifidelity surrogate-based approaches

15:00

Derivative-free multilevel optimization under uncertainty employing higher order moments

15:30

Accurate MLMC estimators for engineering design under uncertainties

14:00

iCal
Olivier Roustant | INSA Toulouse | France

Jhouben Cuesta Ramirez | CEA LETI & Mines Saint-Etienne | France

Sonja Kuhnt | Dortmund University of Applied Sciences and Arts | Germany

Jeremy Rohmer | BRGM | France

MS661: Gaussian process models and metamodels for non Euclidean inputs (Part I of II)

Room:
Interims Lecture Hall 102

Topic:
Surrogate models

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In the last decades there has been renewed interest for Gaussian processes (GP) in statistics and machine learning. New challenges have arisen, especially in uncertainty quantification and optimization for complex systems. The case of continuous inputs has been intensively studied, and can be addressed with existing classes of GPs, such as isotropic (radial) kernels defined with the Euclidean distance. However, numerous applications involve more general non-Euclidean input spaces. This requires the definition of other GPs.

Fortunately, despite the diversity of situations, there are a few common techniques to define valid GPs, such as using a mapping to an Euclidean space. This mini-symposium aims at illustrating the variety of problems encountered along with their specific solutions, as well as the generic techniques. The first part, will focus on the case of discrete inputs in Gaussian process meta-modeling. By discrete input, we mean an input which has a finite number of levels, either ordered or not (it may also be called here “qualitative”, “categorical” or “factor” input). The second part, will present four other cases where the input space can be a permutation, time-varying, a probability distribution or a graph.

14:00

An overview of Gaussian process metamodels with discrete inputs

14:30

Latent variable Bayesian optimization for qualitative and quantitative inputs

15:00

Sensitivity analysis with both continuous and categorical inputs using FANOVA graphs

15:30

Deepening analysis of uncertain categorical inputs using Gaussian processes - application to marine flooding

14:00

iCal
Peter Challenor | University of Exeter | United Kingdom

Ian Vernon | Durham University | United Kingdom

Darren Wilkinson | Newcastle University | United Kingdom

Michael Grosskopf | Los Alamos National Laboratory | United States

MS731: Learning Parameters in Complex Physical Systems with Simulation Experiments (Part I of II)

Room:
Exzellenzzentrum 0003

Topic:
Model error, discrepancy and calibration

Form of presentation:
Mini-symposium

Duration:
120 Minutes

This minisymposium is devoted to recent developments in methodologies, applications, and lessons-learned in estimating physical parameters in complex physical systems. Mathematical models of complex real-world processes have been used to model physical processes of interest in science, engineering, medicine, and business. Computer models (or simulators) often require a set of inputs (some known and specified, others unknown) to generate predictions for physical processes of interest. Physical observations and simulator output allow us to infer both the unknown inputs and the physical process.

Inference about the physical process in the presence of the high-volume output and model uncertainty is challenging, since appropriate uncertainty assessment is the key success to understand the physical process of interest. In the calibration context, the discrepancy between reality and simulators are difficulty to model. In the inverse problem setting, the high-dimensional input space can make the Bayesian inverse computationally challenging.

Bringing selected leading researchers, this minisymposium has been broken into two sessions: calibration (Part I) and inverse problem (Part II). It includes speakers from Europe and North America and is diverse in experience level from fresh PhD graduates to mid-career researchers with backgrounds in statistics, applied mathematics, and engineering. We hope this minisymposium will serve as a nexus to exchange ideas to address this UQ problem.

14:00

History Matching for Physical and Biological Systems

14:30

- CANCELED - Multilevel Emulation and History Matching of EAGLE: an expensive stochastic hydrodynamical Galaxy formation simulation

15:00

Bayesian emulation and calibration of a stochastic computer model geotechnical asset deterioration

15:30

Multifidelity calibration and discrepancy analysis of density functional theory models

16:30

iCal

MT04: Michela Ottobre: Can we make long-term predictions?

Room:
MW HS 2001

Topic:
Prediction

Form of presentation:
Mini-tutorial

Duration:
120 Minutes

16:30

iCal
Eric Parish | Sandia National Labs | United States

Ionut-Gabriel Farcas | Technical University of Munich | Germany

Patrick Blonigan | Sandia National Labs | United States

Christopher Wentland | University of Michigan | United States

MS331: Recent Advances in Reduced-Order Models for Many Query and Time-Critical Problems

Chair(s)
Francesco Rizzi (NexGen Analytics)

Patrick Blonigan (Sandia National Labs)

Kevin Carlberg (Facebook Reality Labs)

Patrick Blonigan (Sandia National Labs)

Kevin Carlberg (Facebook Reality Labs)

Room:
MW HS 0001

Topic:
Reduced order models

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Computational science is a driver of our society’s technological advancement, playing a key role for design, decision making and risk assessment. The ``extreme-scale'' computing era we are living in is enabling a paradigm shift: we no longer approach a problem with a few, target runs for specific choices of parameters and conditions, but we aim increasingly more at combining higher fidelity models with uncertainty quantification (UQ) methods to address, e.g., design optimization and parameter-space exploration. This approach allows us to discover rare events and critical behaviors of a target system, which is key information for high-consequence systems and cutting-edge engineering. If the system of interest is expensive to query, UQ can become impractical to complete within a reasonable amount of time. Reduced-order models (ROMs), due to their accuracy, computational efficiency and certification, constitute a promising technique to overcome this computational barrier, and make high-fidelity predictive simulations feasible for UQ. This mini-symposium aims at presenting recent advances in algorithms, software and applications in the context of reduced-order models and their broad impact for UQ. The talks will cover a broad range of applications, ranging from hypersonics to multiscale flows and plasma microturbulence.

16:30

Windowed Least–Squares Reduced-Order Models for Dynamical Systems

17:00

Context-aware multifidelity Monte Carlo sampling

17:30

- CANCELED - Least-Squares Petrov-Galerkin Reduced-Order Models for Hypersonic Flight Vehicles

18:00

Scalable Closure of Nonlinear Manifold Reduced-Order Models

16:30

iCal
Yanzhao Cao | Auburn University | United States

Tao Zhou | Chinese Academy of Sciences | China

Cosmin Safta | Sandia National Laboratory | United States

Wei Cai | Southern Methodist University | United States

MS321: Data Driven Stochastic Optimization

Room:
MW HS 1801

Topic:
Optimization and optimal control under uncertainty

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Stochastic optimization is an effective approach to solve inverse problems, especially when traditional deterministic optimization methods fail or do not perform well. Important stochastic optimization methods include stochastic gradient descent, Bayesian inference, particle-based Monte Carlo sampling, and many more. With modern data collection techniques, a large amount of data is available as the input in inverse problems, which creates great needs of data driven optimization methods. In this mini-symposium, we focus on discussions of numerical methods related to data driven stochastic optimization and explore applications of data driven stochastic optimization methods in science and engineering.

16:30

Quasi Monte Carlo Stochastic Gradient Descent Method for Optimal Control Problem

17:00

Adaptive multi-fidelity surrogate modeling for Bayesian inference in inverse problems

17:30

Low-Rank Functional Representations for Sensitivity Analysis in Earth System Models

18:00

Algorithms for Wave Scattering of Random Media: A FMM for layered media and a phase shift DNN for high frequency learning

16:30

iCal
Mikael Laaksonen | Lappeenranta-Lahti University of Technology | Finland

Joshua Chen | Oden Institute for Computational Engineering and Sciences, The University of Texas at Austin | United States

Abdellah Chkifa | Mohammed VI Polytechnic University | Morocco

MS032: Kernel, Quasi-Monte Carlo, and Sparse Grid Methods for High-dimensional Approximation and Integration (Part II of III)

Chair(s)
Alexander Gilbert (Universität Heidelberg)

Fabio Nobile (EPFL)

Michael Griebel (University of Bonn and Fraunhofer SCAI)

Yoshihito Kazashi (EPFL)

Fabio Nobile (EPFL)

Michael Griebel (University of Bonn and Fraunhofer SCAI)

Yoshihito Kazashi (EPFL)

Room:
MW HS 0350

Topic:
High-dimensional approximation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

This mini-symposium aims at bringing together people working on kernel and other sampling-based approximation methods for high-dimensional problems, in particular, but not restricted to, quasi-Monte Carlo methods, and sparse grids methods. Kernel methods and the related Gaussian Process surrogate models are a powerful class of numerical methods, and they are often employed in problems arising in uncertainty quantification. Nonetheless, there is much to be explored in their theoretical analysis for UQ applications, which are often formulated as high-dimensional approximation or integration problems.

On the other hand, the theory and applicability of QMC and sparse grid approximation/integration techniques in high or infinite dimensional problems have seen considerable advances in the last years, yet being far from addressing all problems of interest in UQ.

The objective of this mini-symposium is to showcase the late theoretical results and exchange ideas on sampling-based high dimensional integration and approximation methods targeting UQ applications.

16:30

Stochastic collocation method for computing eigenspaces of parameter-dependent operators

17:00

On adaptation of sparse quadrature and sparse polynomial-based Knothe Rosenblatt maps for high-dimensional Bayesian inversion

17:30

On a fast hierarchical sparse grid quadrature and applications

16:30

iCal
Benjamin Sanderse | Centrum Wiskunde & Informatica | Netherlands

Juan Pablo Madrigal Cianci | EPFL | Switzerland

Giacomo Garegnani | EPFL | Switzerland

Mass Per Pettersson | NORCE Norwegian Research Centre AS | Norway

MS781: Adaptive sampling methods for heterogeneous problems

Chair(s)
Sebastian Krumscheid (RWTH Aachen University)

Mass Per Pettersson (NORCE Norwegian Research Centre AS)

Mass Per Pettersson (NORCE Norwegian Research Centre AS)

Room:
MW HS 0250

Topic:
Statistical methods for UQ

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

This minisymposium targets research on adaptive and efficient sampling methods for heterogeneous problems that do not depend smoothly on random (model) parameters. Uncertainty quantification for such problems, especially those that exhibit discontinuities, are notoriously challenging to solve efficiently with existing methods. Due to the lack of regularity usually only sampling based methods remain a robust alternative. However, these methods may converge slowly, unless combined with suitable accelerating techniques such as variance reduction techniques. Even when this is done, such a method’s performance may still be reduced compared to when applied to smooth problems, as demonstrated for instance in the context of multi-level Monte Carlo methods for non-smooth functions [Schwab & Mishra, Krumscheid et al]. The challenges of heterogeneous problems have been demonstrated repeatedly for different classes of methods, including localized generalized polynomial chaos using wavelets [LeMaitre and

Knio] or multi-elements [Wan & Karniadakis]. Common to these methods is that they require a problem dependent adaptation of the sampling procedures in the vicinity of heterogeneous features. Here, adaptivity is to be understood in a wide sense, ranging from machine learning approaches for identifying parameterizations or response functions to discontinuity tracking. In this minisymposium we will discuss techniques that combine such adaptivity with efficient sampling algorithms.

16:30

An adaptive minimum spanning tree method for UQ of discontinuous responses

17:00

A KDE-based Multi-level Markov chain Monte Carlo algorithm

17:30

Model Misspecification And Uncertainty Quantification For Drift Estimation In Multiscale Diffusion Processes

18:00

Adaptive Stratified Sampling for Non-smooth Problems

16:30

iCal
Han Cheng Lie | Universität Potsdam | Germany

Michael Schober | Max Planck Institute for Intelligent Systems | Germany

Filip Tronarp | EEA Aalto University | Finland

Junyang Wang | Newcastle University | United Kingdom

MS022: Probabilistic Numerical Methods for Differential Equations and Linear Algebra (Part II of II)

Chair(s)
Philipp Hennig (University of Tübingen & Max Planck Institute for Intelligent Systems)

Alejandro Diaz (University College London)

Alejandro Diaz (University College London)

Room:
MW HS 1250

Topic:
Probability Theory for UQ

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In many important inverse problems and engineering computations -e.g. numerical weather prediction, medical tomography, reliability analysis- data are related to parameters of interest through the solution of an ordinary or partial differential equation (DE). To proceed with computation, the DE must be discretised and solved through linear algebra methods. However, such discretisation introduces bias into parameter estimates and can in turn cause conclusions to be over-confident. Probabilistic numerical methods for DEs and linear algebra aim to provide uncertainty quantification in the solution space of the DE to properly account for the fact that the governing equations have been altered through discretisation. In contrast to the worst-case error bounds of classical numerical analysis, the stochasticity in DEs and linear solvers serves as the carrier of uncertainty about discretisation error and its impact. This statistical notion of discretisation uncertainty can then be more easily propagated to later inferences, e.g. in a Bayesian inverse problem. Several such probabilistic numerical methods have been developed in recent years, and the connections and distinctions between these methods are starting to be modelled and understood. In particular, an important challenge is to ensure that such uncertainty estimates are well-calibrated. This minisymposium will examine recent advances in both the development and implementation of probabilistic numerical methods in general.

16:30

On the role of exponential integrability of probabilistic integrators for approximate Bayesian inference

17:00

BVPs, computational pipelines and a probabilistic numerics GOODE

17:30

Probabilistic solutions to ordinary differential equations as non-linear Bayesian filtering and smoothing: Gaussian approximations

18:00

Approximate Bayesian solutions to nonlinear differential equations

16:30

iCal
Uwe Ehret | Karlsruhe Institute of Technology | Germany

Mario Teixeira Parente | Technical University of Munich | Germany

Valentina Noacco | University of Bristol | United Kingdom

Xavier Sanchez Vila | Universitat Politècnica de Catalunya | Spain

MS062: Uncertainty Quantification in Hydrology (Part II of II)

Chair(s)
Ana Gonzalez-Nicolas (University of Stuttgart)

Gabriele Chiogna (Technical University of Munich and University of Innsbruck)

Gabriele Chiogna (Technical University of Munich and University of Innsbruck)

Room:
MW ZS 1050

Topic:
Uncertainty propagation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

Hydrological model simulations are often complicated by inevitable uncertainties in initial conditions, boundary conditions, and parameter fields. A proper identification and quantification of such uncertainties are nowadays a must for any modern hydrologist. In this mini symposium, beside presentations focusing on how uncertainty quantification can be properly performed for problems typical of hydrological sciences (e.g., flow and transport in porous media, river and karst spring discharge predictions, surface water-groundwater interaction…), we want to emphasize why uncertainty quantification is relevant in hydrology and its implication for engineering applications.

The minisimposium received funding from the International Graduate School of Science and Engineering of the Technical University of Munich.

16:30

Better informed than uncertain – Information theory as a framework for uncertainty quantification

17:00

Solving a Bayesian Inverse Problem for a Karst Aquifer Model with Active Subspaces

17:30

From uncertainty quantification to uncertainty attribution: what we can learn through global sensitivity analysis and how it can help in the calibration and evaluation of hydrological models

18:00

Model ambiguity in subsurface flow in the presence of limited knowledge in hydraulic conductivities

16:30

iCal
Ahmad Rushdi | Sandia National Laboratories | United States

Jize Zhang | Lawrence Livermore National Laboratory | United States

Theodore Papamarkou | Oak Ridge National Laboratory | United States

Nevin Martin | Sandia National Laboratories | United States

MS472: Leveraging the Interplay Between UQ and ML for Mutual Benefit (Part II of II)

Chair(s)
Justin Newcomer (Sandia National Laboratories)

Gowri Srinivasan (Los Alamos National Laboratory)

Jean-Christophe Weill (CEA)

Gowri Srinivasan (Los Alamos National Laboratory)

Jean-Christophe Weill (CEA)

Room:
MW ZS 2050

Topic:
Physics-informed and data-driven UQ methods

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Machine Learning (ML) has evolved into a core technology in many scientific applications. Solutions often require large labeled datasets to achieve high model accuracy. Unfortunately, this is a major bottleneck for many scientific computing applications, where numerical simulations are very expensive. Training on limited data can lead to significant uncertainties or errors when invoked outside the training space. But the fast execution of ML models once trained also make them ideal for exploring large numbers of runs for Uncertainty Quantification (UQ). Furthermore, many popular ML methods lack the needed mathematical support to prove robustness and reliability to motivate their use in scientific computing and uncertainty quantification UQ applications. This two-part mini-symposium will explore the interplay between ML and UQ, focusing in the following areas: (1) How do we leverage ML successes for scientific computing problems with uncertain inputs? (2) How do we use UQ methods to assess ML predictions and augment them with uncertainty estimates, error bounds, or prediction intervals? Addressing challenges in these areas will lead to greatly improve predictive capabilities. Methods that incorporate mathematical and scientific principles for uncertainty estimates in ML are needed. Literature in statistics can be leveraged for improving the model validation process and advances in UQ and V&V will greatly enhance the mathematical and scientific computing foundations for ML.

16:30

Predictive Uncertainty Estimation in Scientific Machine Learning Models

17:00

Uncertainty quantification of deep learning predictive models: with application to image-based material property prediction

17:30

- CANCELED - Challenges in Bayesian inference via Markov chain Monte Carlo for neural networks

18:00

Credibility Processes for Engineering Analyses Using Machine Learning

16:30

iCal
Thomas Catanach | Sandia National Laboratories | United States

Bobbie Wu | University of Texas at Austin | United States

Tiangang Cui | Monash University | Australia

Nick Winovich | Purdue University | United States

MS591: Bayesian inversions with computationally expensive models

Room:
MW ZS 1450

Topic:
Inverse problems

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Scientific and engineering models, which are generally partial differential equations (PDEs), often contain uncertain model parameters, initial conditions, and boundary conditions. These are often inferred by fitting to experimental or field observables. Bayesian inverse problems allow these unknowns to be modeled as random variables or fields and estimate a probability density for them. This is usually done via sampling, e.g., via a Markov chain Monte Carlo sampler. The probability density captures the uncertainty in the estimated quantities due to shortcomings of the model and sparsity of the data.

Computationally expensive PDE simulators do not allow their direct employment in samplers, and we often take recourse to statistical emulators. Training data for the emulators can be difficult to generate. We either reduce the dimensionality beforehand, or take recourse to sparse-grid sampling. Priors are generally known only as bounds, but arbitrary parameter combinations sampled from the resulting multidimensional uniform distributions may not be physically realistic and the PDE simulator may not even run.

We invite talks in dimensionality reduction, the construction of computationally inexpensive proxies of scientific/engineering simulators, strategies to fashion a physically realistic prior and other practical methods required to solve inverse problems of engineering/scientific interest. Examples where such methods have been used to solve inverse problems are also welcome.

16:30

Integrating Multifidelity Models into Sequential Tempered Markov Chain Monte Carlo

17:00

- CANCELED - Bayesian Inference and Optimal Experimental Design for System Identification of Material Physics Phenomena

17:30

- CANCELED - Bayesian filtering and parameter estimation without particles -- a tensor-based approach

18:00

VoroSpokes Sampling with Applications to Bayesian Inference

16:30

iCal
Derek Posselt | Jet Propulsion Laboratory, California Institute of Technology | United States

Tijana Janjic Pfander | University of Munich | Germany

Sabrina Sanchez | Max Planck Institute for Solar System Research | Germany

Yvonne Ruckstuhl | University of Munich | Germany

MS792: Bayesian Inference in Earth Science (Part II of II)

Chair(s)
Spencer Lunderman (University of Arizona)

Derek Posselt (Jet Propulsion Laboratory, California Institute of Technology)

Matthias Morzfeld (University of Arizona)

Derek Posselt (Jet Propulsion Laboratory, California Institute of Technology)

Matthias Morzfeld (University of Arizona)

Room:
MW ZS 1550

Topic:
Data assimilation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

There is typically a mismatch between observations of a process, and its representation in a mathematical or numerical model. Such error arises because the model is incomplete or approximate, and errors are amplified by noise in the observations, as well as uncertain, or completely unknown, model states and parameters. In Earth science, errors of these types must be quantified, and a natural tool to do so is Bayesian inference, where errors are described via conditional probabilities defined for the model, its parameters, and the observations. This mini-symposium will focus on the numerical solution of Bayesian inference problems in Earth sciences which are usually characterized by a large dimension (many parameters and states) and few observations (relative to the number of states and parameters). Moreover, Earth science applications require solutions to three types of Bayesian inference problems: state estimation (data assimilation), parameter estimation, and joint state and parameter estimation. Our mini-symposium will showcase Bayesian inference "in action" in Earth science. It will provide an opportunity for interaction among applied mathematicians, interested in the numerics of Bayesian inference, and Earth scientists, who use Bayesian inference to break new ground in their respective fields.

16:30

- CANCELED - Bayesian Inference for Cloud and Precipitation Model Parameter Estimation

17:00

Data assimilation on convective scale with positivity preservation

17:30

Geomagnetic data assimilation, a window to the Earth’s core dynamics

18:00

- NEW - Parameter and state estimation with ensemble Kalman filter based algorithms for convective-scale applications

16:30

iCal
Benjamin Peherstorfer | Courant Institute, New York University | United States

Robert Scheichl | Ruprecht-Karls University Heidelberg | Germany

Ian Langmore | Google | United States

Sergey Dolgov | University of Bath | United Kingdom

MS112: Inference and preconditioning via Stein methods, flows, and other transport maps (Part II of II)

Chair(s)
Benjamin Peherstorfer (Courant Institute, New York University)

Youssef Marzouk (Massachusetts Institute of Technology)

Yaoliang Yu (University of Waterloo)

Youssef Marzouk (Massachusetts Institute of Technology)

Yaoliang Yu (University of Waterloo)

Room:
MW 2250

Topic:
Statistical methods for UQ

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

Transport maps are deterministic couplings between probability measures with broad applications in uncertainty quantification and machine learning. They have been used for posterior sampling in Bayesian inference, for accelerating Markov chain Monte Carlo and importance sampling algorithms, and as building blocks of generative models and density estimation methods. More broadly, transport---including but not limited to optimal transport---provides an important mathematical foundation for many tools in machine learning and uncertainty quantification. The recent surge of interest in transport maps has been accompanied by efficient numerical methods that make constructing and learning such maps tractable in high dimensions and for large data sets. This minisymposium brings together researchers from uncertainty quantification and machine learning to discuss recent advances in theory, numerics, and applications of transport maps and related techniques.

16:30

A transport-based multifidelity preconditioner for Markov chain Monte Carlo

17:00

HINT: Hierarchical Invertible Neural Transport for General and Sequential Bayesian inference

17:30

Preconditioning at scale for a fusion plasma inverse problem

18:00

- MOVED from MS192 - Deep tensor product Rosenblatt transformation for sampling of high-dimensional distributions

16:30

iCal
Sebastian Mitusch | Simula Research Laboratory | Norway

Katherine Johnston | Sandia National Laboratories | United States

Matthew Parno | US Army Cold Regions Research and Engineering Laboratory (CRREL) | United States

Umberto Villa | Washington University in St. Louis | United States

MS621: Computational tools for inverse problems governed by PDEs and UQ

Chair(s)
Noemi Petra (University of California, Merced)

Matthew Parno (US Army Cold Regions Research and Engineering Laboratory )

Umberto Villa (Washington University in St. Louis)

Matthew Parno (US Army Cold Regions Research and Engineering Laboratory )

Umberto Villa (Washington University in St. Louis)

Room:
MW 1701

Topic:
Inverse problems

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In recent years Bayesian inference has emerged as the most comprehensive and systematic framework for formulating and solving inverse problems with quantified uncertainties. However, the solution of Bayesian inverse problems governed by PDEs is extremely challenging; complex forward models or large parameter dimensions can make Bayesian inversion prohibitive with standard methods. In addition, often one has to solve a PDE-constrained optimization subproblem several times. Recent years have seen intensive efforts to develop advanced algorithms aimed at this class of problems; however, due to the complexity of the algorithms and potential dependencies on derivative information, they have remained buried in the literature and out of the reach of a broad community of scientists and engineers who solve inverse problems. The goal of this minisymposium is to present software frameworks that make advanced algorithms more accessible to domain scientists and provide an environment that expedites the development of new algorithms. These software frameworks can also be used as teaching tools that can be used to educate researchers and practitioners who are new to inverse problems, PDE-constrained optimization, the Bayesian inference framework and UQ in general.

16:30

dolfin-adjoint: A Python framework for automated adjoints of PDEs

17:00

Bayesian Inference Tools in the UQTk UQ Toolkit

17:30

Using the MIT Uncertainty Quantification (MUQ) library for high-dimensional inverse problems

18:00

End-to-end Uncertainty Quantification with hIPPYlib

16:30

iCal
Andrea Scarinci | Massachusets Institute of Technology (MIT) | United States

Mélanie C. Rochoux | CERFACS | France

Kjetil Lye | ETH Zurich | Switzerland

Alban Farchi | CEREA (Ecole des Ponts ParisTech and EDF R&D) | France

MS252: Similarity measures and distances in forward and inverse UQ problems (Part II of II)

Chair(s)
Didier Lucor (LIMSI - CNRS)

Mohammad Motamed (The University of New Mexico)

Lionel Mathelin (LIMSI - CNRS)

Mohammad Motamed (The University of New Mexico)

Lionel Mathelin (LIMSI - CNRS)

Room:
MW HS 0337

Topic:
Other

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Organized in co-operation with

GAMM AG UQ

[ Moved from MW HS 2235 ]

The analysis and comparison of dynamic objects and deforming shapes is important in many real-world applications. Examples include wildfire front-tracking problems, impulse propagation in cardiac tissues, tumor growth, oil reservoir and spill simulations, and pollutant plume dispersion, just to name a few. There are several difficulties that can make the analysis a daunting task and hence need to be addressed: 1) the problem is subjected to uncertainty in the location of structures due to numerical errors, measurement noise, and/or intrinsic variations in the system; 2) strong shape deformations and topological changes may not be well captured at all scales; and 3) the notion of distance or similarity between objects can be characterized in various ways.

This situation has fostered a recent body of work focused on both analytical and computational developments in metric spaces. As an example, the Wasserstein metric has become an increasingly popular tool in such diverse fields as image processing, optimization, neural networks, seismic imaging, and numerical conservation laws. It opens up promising avenues for uncertainty quantification, Bayesian inference and data assimilation, where robust comparisons and mappings between different probability measures are often needed.

This MS will review recent advances, applications and remaining challenges of tailored metric spaces and similarity measures for structure-sensitive uncertainty quantification and inference problems.

16:30

Model misspecification and transport-Lagrangian distances in seismic inversion

17:00

Sensitivity Analysis in Wildland Fire Modeling for Front Data Assimilation

17:30

Wasserstein Convergence of Finite Volume Schemes for Hyperbolic Conservation Laws

18:00

Using the Wasserstein distance to compare fields of pollutants: application to the radionuclide atmospheric dispersion of the Fukushima-Daiichi accident

16:30

iCal
Samuel Rudy | Massachusetss Insitute of Technology | United States

Shanyin Tong | NYU Courant | United States

Hendrik Dijkstra | Utrecht University | Netherlands

Vera Melinda Galfi | University of Hamburg | Germany

MS102: Statistical Prediction and Quantification of Extreme Events in Complex Systems (Part II of II)

Chair(s)
Antoine Blanchard (Massachusetts Institute of Technology)

Themistoklis P. Sapsis (Massachusetts Institute of Technology)

Themistoklis P. Sapsis (Massachusetts Institute of Technology)

Room:
IAS 0.001

Topic:
Rare events and Risk

Form of presentation:
Mini-symposium

Duration:
120 Minutes

Extreme events are short-lived episodes occurring due to exogenous causes or internal instabilities during which observables significantly depart from their mean values. A great deal of effort has been devoted to predicting and statistically quantifying extreme events because they can have catastrophic consequences (e.g., structural failure, rogue waves, extreme weather conditions, and market crashes). This is an arduous task because the systems that give rise to extreme events are most often highly complex and strongly nonlinear. This mini-symposium provides a venue to review the latest advances in the field.

16:30

- NEW - Sparse Methods for Bayesian Linear Regression

17:00

- CANCELED - Estimation of Extreme Tsunami Waves Using Large Deviation Theory

17:30

Transition Probabilities of Noise-Induced Transitions of the Atlantic Ocean Circulation

18:00

A Large Deviation Theory-Based Analysis of Heat Waves and Cold Spells in a Simplified Model of the General Circulation of the Atmosphere

16:30

iCal
Emil Constantinescu | Argonne National Laboratory | United States

Thordis Thorarinsdottir | Norwegian Computing Center | Norway

Petra Friederichs | University of Bonn | Germany

Jochen Broecker | University of Reading | United Kingdom

MS291: Statistical evaluation and scoring of complex simulations and predictions

Chair(s)
Emil Constantinescu (Argonne National Laboratory)

Julie Bessac (Argonne National Laboratory)

Julie Bessac (Argonne National Laboratory)

Room:
IAS 4.001

Topic:
Statistical methods for UQ

Form of presentation:
Mini-symposium

Duration:
120 Minutes

A key challenge associated with simulations and predictions of complex systems is to evaluate the quality of these datasets and the ability of the underlying model to reproduce physically relevant simulations. In statistics one way to quantitatively evaluate and rank models is statistical scoring. This is typically based on scalar metrics and takes as input verification data and output from the model to be evaluated. While evaluating model simulations or predictions, one aims to detect bias, trends, outliers, or correlation misspecification. Methods to evaluate the quality of unidimensional outputs are well established; however, issues remains related to score approximation and uncertainty. Additionally, the evaluation of multidimensional outputs or ensemble of outputs has been addressed in the literature relatively recently and remains challenging. We will discuss these challenges associated with evaluating unidimensional and multidimensional simulations or predictions.

16:30

Using scoring rules to solve stochastic inverse problems

17:00

Generating proper scoring rules for high-dimensional objects using summary statistics

17:30

Using wavelets to verify the scale structure of precipitation forecasts

18:00

Evaluating reliability of forecasting systems under serial correlation

16:30

iCal
Hillary Fairbanks | Lawrence Livermore National Laboratory | United States

Nicholas Galioto | University of Michigan | United States

Sam Friedman | Texas A&M University | United States

Gianluca Geraci | Sandia National Laboratories | United States

MS683: Multilevel and Multifidelity approaches for forward/inverse Uncertainty Quantification and optimization under uncertainty (Part III of III)

Chair(s)
Panagiotis Tsilifis (EPFL)

Gianluca Geraci (Sandia National Laboratories)

Alex Gorodetsky (University of Michigan)

John Jakeman (Sandia National Laboratories)

Juan Pablo Madrigal Cianci (EPFL)

Michael Eldred (Sandia National Laboratories)

Gianluca Geraci (Sandia National Laboratories)

Alex Gorodetsky (University of Michigan)

John Jakeman (Sandia National Laboratories)

Juan Pablo Madrigal Cianci (EPFL)

Michael Eldred (Sandia National Laboratories)

Room:
Interims Lecture Hall 101

Topic:
Uncertainty propagation

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In the last decades, the advancements in both computer hardware/architectures

and algorithms enabled numerical simulations at unprecedented scales. In parallel,

Uncertainty Quantification (UQ) evolved as a crucial task to enable predictive

numerical simulations. Therefore, a great effort has been devoted in advancing the UQ algorithms

in order to enable UQ for expensive numerical simulations, however the combination of an extremely

large computational cost associated to the evaluation of a high-fidelity model and the presence of a moderate/large

set of uncertainty parameters (often correlated to the complexity of the numerical/physical assumptions)

still represents a formidable challenge for UQ.

Multilevel and multifidelity strategies have been introduced to circumvent these difficulties by

reducing the computational cost required to perform UQ with high-fidelity simulations. The

main idea is to optimally combine simulations of increasingly resolution levels or model fidelities

in order to control the overall accuracy of the surrogates/estimators. This task is accomplished by

combining large number of less accurate numerical simulations with only a limited number of high-fidelity,

numerically expensive, code realizations. In this minisymposium we present contributions related to the state-of-the-art in both forward and inverse multilevel/multifidelity UQ and related areas as optimization under uncertainty.

16:30

A PDE-Based Hierarchical Sampling Approach for Inputs to Multilevel Markov Chain Monte Carlo

17:00

Approximate control variate approaches for Bayesian inverse problems

17:30

Mufti-fidelity uncertainty quantification for coupled systems

18:00

Recent advancements in sampling approaches for multifidleity uncertainty quantification

16:30

iCal
Jean-Michel Loubès | Institut de Mathématiques de Toulouse | France

Ali Hebbal | ONERA & Université de Lille | France

François Bachoc | Institut de Mathématiques de Toulouse | France

Nicolas Durrande | Prowler.io | United Kingdom

MS662: Gaussian process models and metamodels for non Euclidean inputs (Part II of II)

Room:
Interims Lecture Hall 102

Topic:
Surrogate models

Form of presentation:
Mini-symposium

Duration:
120 Minutes

In the last decades there has been renewed interest for Gaussian processes (GP) in statistics and machine learning. New challenges have arisen, especially in uncertainty quantification and optimization for complex systems. The case of continuous inputs has been intensively studied, and can be addressed with existing classes of GPs, such as isotropic (radial) kernels defined with the Euclidean distance. However, numerous applications involve more general non-Euclidean input spaces. This requires the definition of other GPs.

Fortunately, despite the diversity of situations, there are a few common techniques to define valid GPs, such as using a mapping to an Euclidean space. This mini-symposium aims at illustrating the variety of problems encountered along with their specific solutions, as well as the generic techniques. The first part, will focus on the case of discrete inputs in Gaussian process meta-modeling. By discrete input, we mean an input which has a finite number of levels, either ordered or not (it may also be called here “qualitative”, “categorical” or “factor” input). The second part, will present four other cases where the input space can be a permutation, time-varying, a probability distribution or a graph.

16:30

Gaussian processes on distributions

17:00

Multi-Fidelity modeling with varying input space dimensions using Deep Gaussian Processes

17:30

Gaussian processes indexed on the symmetric group: prediction and learning

18:00

Variational inference for Gaussian Markov Random Fields

16:30

iCal
Robert Gramacy | Virginia Tech | United States

Babak Maboudi Afkham | Stuttgart University | Germany

Pranjal Pranjal | Virginia Tech | United States

MS732: Learning Parameters in Complex Physical Systems with Simulation Experiments (Part II of II)

Room:
Exzellenzzentrum 0003

Topic:
Model error, discrepancy and calibration

Form of presentation:
Mini-symposium

Duration:
120 Minutes

This minisymposium is devoted to recent developments in methodologies, applications, and lessons-learned in estimating physical parameters in complex physical systems. Mathematical models of complex real-world processes have been used to model physical processes of interest in science, engineering, medicine, and business. Computer models (or simulators) often require a set of inputs (some known and specified, others unknown) to generate predictions for physical processes of interest. Physical observations and simulator output allow us to infer both the unknown inputs and the physical process.

Inference about the physical process in the presence of the high-volume output and model uncertainty is challenging, since appropriate uncertainty assessment is the key success to understand the physical process of interest. In the calibration context, the discrepancy between reality and simulators are difficulty to model. In the inverse problem setting, the high-dimensional input space can make the Bayesian inverse computationally challenging.

Bringing selected leading researchers, this minisymposium has been broken into two sessions: calibration (Part I) and inverse problem (Part II). It includes speakers from Europe and North America and is diverse in experience level from fresh PhD graduates to mid-career researchers with backgrounds in statistics, applied mathematics, and engineering. We hope this minisymposium will serve as a nexus to exchange ideas to address this UQ problem.

17:00

Replication or exploration? Active learning and inversion for stochastic simulation experiments

17:30

- MOVED FROM CT01 - Detecting jumps in a jump-discontinuous random field using deep neural networks

18:00

Efficient Bayesian inversion for UQ for high dimensional inverse problems