This workshop will be offered virtually. The in-person meeting has been cancelled due to the COVID-19 outbreak. A schedule for virtual talks is posted below. Accepted participants will be notified how to access the virtual presentations. All other interested parties can view the talks via live-stream.
Organizing Committee
Abstract

The advancement in computing and storage capabilities of modern computational clusters fosters use of novel statistical techniques in machine learning and deep networks. Such data-driven techniques allow one to learn model features and characteristics that are difficult for mathematical methods alone to reveal. Many computational methods achieve model and complexity discovery using methods that lie at the nexus of mathematical, statistical, and computational disciplines. Statistical methods often employ “big data” approaches that glean predictive capability from diverse and enormous databases of information. Emerging methods in machine learning and deep networks can provide impressive results. This workshop gathers researchers at the frontier of large-scale statistical computation, data science, tensor decompositions and approximations, and data-driven model learning, to focus on modern challenges that use data to reduce complexity of models.

Image for "VIRTUAL ONLY: Computational Statistics and Data-Driven Models"

Confirmed Speakers & Participants

Talks will be presented virtually or in-person as indicated in the schedule below.

  • Speaker
  • Poster Presenter
  • Attendee
  • Virtual Attendee

Workshop Schedule

Monday, April 20, 2020
TimeEventLocationMaterials
9:50 - 10:00am EDTWelcome - ICERM Director  
10:00 - 10:45am EDTFast Discovery of Pairwise Interactions in High Dimensions using Bayes - Tamara Broderick, MIT 
11:00 - 11:15am EDTCoffee/Tea Break  
11:15 - 12:00pm EDTLearning Interaction laws in particle- and agent-based systems - Mauro Maggioni, Johns Hopkins University 
12:15 - 1:30pm EDTBreak for Lunch / Free Time  
1:30 - 2:15pm EDTFrom Fourier to Koopman - Spectral Methods for Long-term Time Series Prediction - Nathan Kutz, University of Washington 
2:30 - 2:45pm EDTCoffee/Tea Break  
2:45 - 3:30pm EDTKernel Analog Forecasting for Multiscale Problems - Krithika Manohar, California Institute of Technology 
3:45 - 4:30pm EDTUsing dynamical systems ideas to combine in a principled way data-driven models and domain-driven models - Michael Mahoney, ICSI and Department of Statistics, UC Berkeley 
4:30 - 4:45pm EDTClose the day  
Tuesday, April 21, 2020
TimeEventLocationMaterials
10:00 - 10:45am EDTUnderstanding and mitigating gradient flow pathologies in physics-informed neural networks - Paris Perdikaris, University of Pennsylvania 
11:00 - 11:15am EDTCoffee/Tea Break  
11:15 - 12:00pm EDTDeep Generative Models for Scientific Applications - Jeffrey Regier, University of Michigan 
12:15 - 1:30pm EDTBreak for Lunch / Free Time  
1:30 - 2:15pm EDTCan graph neural networks count substructures? - Soledad Villar, New York University 
2:30 - 2:45pm EDTCoffee/Tea Break  
2:45 - 3:30pm EDTMetric representations - Algorithms and Geometry - Anna Gilbert, University of Michigan 
3:45 - 4:00pm EDTClose the day  
Wednesday, April 22, 2020
TimeEventLocationMaterials
10:00 - 10:45am EDTOptimal experimental design for the quantification of model uncertainty - Karen Veroy-Grepl, Eindhoven University of Technology 
11:00 - 11:15am EDTCoffee/Tea Break  
11:15 - 12:00pm EDTRandomized Kaczmarz for Tensor Systems - Anna Ma, UCI 
12:15 - 1:30pm EDTBreak for Lunch / Free Time  
1:30 - 2:15pm EDTModeling dynamical systems from data - Serkan Gugercin, Virginia Tech 
2:30 - 2:45pm EDTCoffee/Tea Break  
2:45 - 3:30pm EDTReduced Order Model Approach to Inverse Scattering - Jorn Zimmerling, University of Michigan 
3:45 - 4:00pm EDTClose the day  
Thursday, April 23, 2020
TimeEventLocationMaterials
10:00 - 10:45am EDTTransport and Multilevel Approaches for Large-Scale PDE-Constrained Bayesian Inference - Robert Scheichl, Heidelberg University 
11:00 - 11:15am EDTCoffee/Tea Break  
11:15 - 12:00pm EDTOutput-weighted optimal sampling for Bayesian regression and rare event statistics using few samples - Themis Sapsis , MIT 
12:15 - 1:30pm EDTBreak for Lunch / Free Time  
1:30 - 2:15pm EDTExtreme event probability estimation using large-deviation theory and PDE-constrained optimization, with application to tsunamis - Georg Stadler, CIMS NYU 
2:30 - 2:45pm EDTCoffee/Tea Break  
2:45 - 3:30pm EDTCharacterizing Markov process dynamics from trajectory data - Jonathan Weare, Courant Institute, NYU 
3:45 - 4:30pm EDTData-driven methods for modeling and control of complex biological systems - Marcella Gomez, UC Santa Cruz 
4:30 - 4:45pm EDTClose the day  
Friday, April 24, 2020
TimeEventLocationMaterials
10:00 - 10:45am EDTHistory Matching for Inverse Modelling in Physical and Biological Systems - Peter Challenor, University of Exeter 
11:00 - 11:15am EDTCoffee/Tea Break  
11:15 - 12:00pm EDTData-Driven Mechanistic Models -- Design and Inference - Babak Shahbaba, UC Irvine 
12:15 - 1:30pm EDTBreak for Lunch / Free Time  
1:30 - 2:15pm EDTOperator-split reduced order models for problems with moving interfaces. - George Biros, ICES and University of Texas at Austin 
2:30 - 3:00pm EDTClose the workshop  

Poster Session Gallery

Learning context-aware surrogate models for multifidelity importance sampling and Bayesian inverse problems

Terrence Alsup

Differentiating between single particle transport types

Keisha Cook

Fast and Accurate Algorithms for Cosmic Microwave Background Radiation Data on HEALPix Points

Kathryn Drake

emgr - EMpirical GRamian Framework

Christian Himpe

Incorporating physical constraints in deep probabilistic models of coarse-grained dynamics

Sebastian Kaltenbach

System-Theoretic Model Reduction in pyMOR

Petar Mlinarić

Density Tracking by Quadrature for High Dimensional Stochastic Differential Equations

Ryleigh Moore

Machine Learning and Data-Driven Approaches in Spatial Statistics

Sarah Soleiman

Data Driven Models for Solving Partial Differential Equations

Thomas Torku

Analysis of the Ratio of L1 and L2 Norms in Compressed Sensing

Yiming Xu

Hinčin's theorem for additive free convolutions of R-diagonal *-distributions

Cong Zhou

Associated Semester Workshops

Mathematics of Reduced Order Models
Image for "Mathematics of Reduced Order Models "
Feb 17 - 21, 2020
Semester Workshop

Lecture Videos

Operator-split reduced order models for problems with moving interfaces

George Biros
ICES
April 24, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract

Data-Driven Mechanistic Models -- Design and Inference

Babak Shahbaba
University of California - Irvine
April 24, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract

History Matching for Inverse Modelling in Physical and Biological Systems

Peter Challenor
University of Exeter
April 24, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Characterizing Markov process dynamics from trajectory data

Jonathan Weare
Courant Institute, New York University
April 23, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Slides

Extreme event probability estimation using large-deviation theory and PDE-constrained optimization, with application to tsunamis

Georg Stadler
CIMS NYU
April 23, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract

Output-weighted optimal sampling for Bayesian regression and rare event statistics using few samples

Themistoklis Sapsis
MIT
April 23, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Transport and Multilevel Approaches for Large-Scale PDE-Constrained Bayesian Inference

Robert Scheichl
University of Heidelberg
April 23, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Reduced Order Model Approach to Inverse Scattering

Jorn Zimmerling
University of Michigan
April 22, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Modeling dynamical systems from data

Serkan Gugercin
Virginia Tech
April 22, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract

Optimal experimental design for the quantification of model uncertainty

Karen Veroy-Grepl
TU Eindhoven
April 22, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract

Metric representations - Algorithms and Geometry

Anna Gilbert
University of Michigan
April 21, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Understanding and mitigating gradient flow pathologies in physics-informed neural networks

Paris Perdikaris
University of Pennsylvania
April 21, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Kernel Analog Forecasting for Multiscale Problems

Krithika Manohar
California Institute of Technology
April 20, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

From Fourier to Koopman - Spectral Methods for Long-term Time Series Prediction

J. Nathan Kutz
University of Washington
April 20, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides

Welcome and Fast Discovery of Pairwise Interactions in High Dimensions using Bayes

Tamara Broderick
Massachusetts Institute of Technology
April 20, 2020
VIRTUAL ONLY: Computational Statistics and Data-Driven Models
Play     Abstract     Slides