Skip to the content.

IAMCS Workshop in Large-Scale Inverse Problems and Uncertainty Quantification

Description

The ability to solve large-scale inverse problems and to quantify uncertainty is an essential precondition towards the goal of accurate predictions in many data-intensive applications. Challenges in this area are numerous due to multiple scales and uncertainties in physical processes. The workshop will focus on:

  • Efficient data assimilation techniques
  • Robust and accurate model reduction techniques
  • Constructing and assessing the statistical accuracy of emulators
  • Implementation of these methods on today's computer architectures

This workshop brings together researchers working in these areas, with the goal of developing and improving systematic techniques for large-scale inverse problems and uncertainty quantification. It will consist of a number of plenary and invited presentations. We particularly encourage poster presentations.

The workshop will be hosted by Texas A&M University in College Station, Texas, and is mainly supported by the Institute for Applied Mathematics and Computational Science (IAMCS). The workshop is also supported by the Institute for Scientific Computation. There is no registration fee. Interested participants should contact inverseuq@gmail.com. Information will be provided at http://isc.tamu.edu/events/inverseuq/ and will be updated frequently.

Organizing Committee

  • Wolfgang Bangerth, Texas A&M University
  • Akhil Datta-Gupta, Texas A&M University
  • Yalchin Efendiev, Texas A&M University
  • Bani Mallick (Chair), Texas A&M University

Return to the Quick Links Selection Button.

Hotel and Accommodations

Hotel accommodations are available for our guests at a rate of $93.00 plus 15% tax for the nights of and with:

Hyatt Place College Station
1100 University Drive East, College Station, TX 77840
Phone: +1.979.846.9800
Group Code: INSA

Reservations can be made either via telephone at 1.888.HYATT.HP (1.888.492.8847) or online at http://www.hyatt.com. If making a reservation by telephone, be sure to mention the group code INSA in order to receive the negotiated discount rate. Likewise, if making a reservation online, you must input the group code INSA into the special code area in order to see the group discounted rate.

In order to ensure room availability, you must book with Hyatt Place College Station by using the group code INSA. There is a limited number of rooms available so early booking is strongly encouraged. After , or until all reserved rooms are sold-out, the group code will no longer be valid and you may need to reserve a room elsewhere.

More information about the College Station area is available online at http://www.visitaggieland.com/.

Return to the Quick Links Selection Button.

Speakers

Plenary Speakers

George Biros, Georgia Institute of Technology
A Fast Algorithm for the Inverse Medium Problem with Multiple Sources
Geir Evensen, Statoil Research Centre
The Use of Ensemble Methods for History Matching
David Higdon, Los Alamos National Laboratory
Emulating the Nonlinear Matter Power Spectrum for the Universe
Nicholas Zabaras, Cornell University
Model Reduction of Stochastic Systems in Random Heterogeneous Media

Invited Speakers

Carsten Burstedde, University of Texas at Austin
Scalable Algorithms for Large-Scale Inverse Problems Under Uncertainty
Yuguang Chen, Chevron Energy Technology Company
Uncertainty Quantification for Subsurface Flow Problems Using Coarse-Scale Models
Bruce Fryxell, University of Michigan
Calibration and Prediction in a Radiative Shock Experiment
Eduardo Gildin, Texas A&M University
Model Order Reduction in Porous Media Flow
Behnam Jafarpour, Texas A&M University
Model Error Identification Using Sparsity-Based Inversion Techniques for Subsurface Characterization
Hector Klie, ConocoPhillips Company
Dealing with Uncertainties and the Curse of Dimensionality in Closed-Loop Reservoir Management
Stein Krogstad, SINTEF
Local POD-Based Multiscale Mixed FEMs for Model Reduction of Multiphase Compressible Flow
Cornelia Kronsbein, Fraunhofer ITWM (Kaiserslautern, Germany)
Multi-Level Monte Carlo for Multi-Scale Simulations of Flows in Heterogeneous Media
Faming Liang, Texas A&M University
A Dynamically Weight Particle Filter for Sea Surface Temperature Prediction
Guang Lin, Pacific Northwest National Laboratory
Uncertainty Quantification & Dynamic State Estimation of Power Grid System
Marko Maucec, Halliburton/Landmark Graphics Corporation
Towards Geology-Guided Quantitative Uncertainty Management in Dynamic Model Inversion
Luis Tenorio, Colorado School of Mines
Data Analysis Tools for Uncertainty Quantification of Inverse Problems
Xiao-Hui Wu, ExxonMobil
Multilevel Methods for Uncertainty Quantification - Opportunities and Challenges

Return to the Quick Links Selection Button.

Schedule

All workshop presentations will take place in the Stephen W. Hawking Auditorium of the George P. and Cynthia Woods Mitchell Institute for Fundamental Physics and Astronomy (MIST). The poster session will take place in the MIST foyer. The formal dinner will be held in The University Club located in Rudder Tower. Please note that the schedule is tentative and subject to change. All times listed are local time.

Day 1:

Opening Remarks
Geir Evensen, Statoil Research Centre
The Use of Ensemble Methods for History Matching
Behnam Jafarpour, Texas A&M University
Model Error Identification Using Sparsity-Based Inversion Techniques for Subsurface Characterization
Guang Lin, Pacific Northwest National Laboratory
Uncertainty Quantification & Dynamic State Estimation of Power Grid System
Break
Xiao-Hui Wu, ExxonMobil
Multilevel Methods for Uncertainty Quantification - Opportunities and Challenges
Hector Klie, ConocoPhillips Company
Dealing with Uncertainties and the Curse of Dimensionality in Closed-Loop Reservoir Management
George Biros, Georgia Institute of Technology
A Fast Algorithm for the Inverse Medium Problem with Multiple Sources
Lunch Break
Nicholas Zabaras, Cornell University
Model Reduction of Stochastic Systems in Random Heterogeneous Media
Stein Krogstad, SINTEF
Local POD-Based Multiscale Mixed FEMs for Model Reduction of Multiphase Compressible Flow
Marko Maucec, Halliburton/Landmark Graphics Corporation
Towards Geology-Guided Quantitative Uncertainty Management in Dynamic Model Inversion
Yuguang Chen, Chevron Energy Technology Company
Uncertainty Quantification for Subsurface Flow Problems Using Coarse-Scale Models
Poster Session
George P. and Cynthia Woods Mitchell Institute for Fundamental Physics and Astronomy (MIST) Foyer
Formal Dinner
The University Club
Rudder Tower

Day 2:

David Higdon, Los Alamos National Laboratory
Emulating the Nonlinear Matter Power Spectrum for the Universe
Luis Tenorio, Colorado School of Mines
Data Analysis Tools for Uncertainty Quantification of Inverse Problems
Bruce Fryxell, University of Michigan
Calibration and Prediction in a Radiative Shock Experiment
Break
Carsten Burstedde, University of Texas at Austin
Scalable Algorithms for Large-Scale Inverse Problems Under Uncertainty
Eduardo Gildin, Texas A&M University
Model Order Reduction in Porous Media Flow
Break
Faming Liang, Texas A&M University
A Dynamically Weight Particle Filter for Sea Surface Temperature Prediction
Cornelia Kronsbein, Fraunhofer ITWM (Kaiserslautern, Germany)
Multi-Level Monte Carlo for Multi-Scale Simulations of Flows in Heterogeneous Media
Closing Remarks
After
Finite Element Rodeo

Return to the Quick Links Selection Button.

Contact Information

If you have any questions concerning this workshop, email inverseuq@gmail.com.

Return to the Quick Links Selection Button.

Finite Element Rodeo

The finite element rodeo is an annual, informal conference on finite element methods that rotates between several universities in Texas and Louisiana. In the beginning, we ask for all those willing to speak to come forward. We will use this information to determine the length of the talks and arrange a schedule so that the total available time is then divided by the number of speakers. Usually, the time alloted per talk is 10–15 minutes. Talks by graduate students and postdocs are particularly encouraged.

The edition of the traditional finite element rodeo will be held at Texas A&M on starting at . Registration starts at , and all participants need to be registered before

More information can be found by visiting http://www.math.tamu.edu/~bonito/Workshops/2011_FEMRodeo/femrodeo_2011.php.

Return to the Quick Links Selection Button.