8:30-8:45 a.m. — Welcome | Tim Germann, Arvind Mohan & Gowri Srinivasan, Los Alamos National Laboratory

8:45 a.m.-Noon – Session 1 | ML for Molecular Energies

8:45-9:30 a.m. – Justin Smith, Los Alamos National Laboratory | Accelerated modeling of atomistic physics with machine learning

9:30-10 a.m. – Jean-Bernard Maillet, CEA | "Machine Learning" for interatomic potentials: examples and applications

10-10:20 a.m. — Break

10:20-10:40 a.m. – Nicholas Lubbers, Los Alamos National Laboratory | Realization of Physical Principles in Atomistic Machine Learning

10:40-11 a.m. – Ying Shi Teh, CalTech | Accelerating electronic structure calculations with machine learning

11-11:20 a.m. – Jiaxin Zhang, Oak Ridge National Laboratory | Deep Learning Method to Accelerate Monte Carlo Simulation

11:20-11:40 a.m. – Michael Grosskopf, Los Alamos National Laboratory | Using machine learning to identify sources of bias in simulation of nuclear data validation benchmarks

11:40-Noon – Discussions

Noon-1:45 p.m. — Lunch 1:45-5:15 p.m. – Session 2 | ML for Interatomic Potentials and Fluid Dynamics

1:45-2:30 p.m. – Noam Bernstein, Naval Research Laboratory | Approximating a potential energy surface: machine learning for interatomic potentials

2:30-3 p.m. – Aidan Thompson, Sandia National Laboratories | Predictive Atomistic Simulations of Materials using SNAP Machine-Learning Interatomic Potentials

3-3:20 p.m. – Break

3:20-3:40 p.m. – Ben Nebgen, Los Alamos National Laboratory | Deep Neural Networks for Effective Hamiltonian Parameterization

3:40-4 p.m. – Mingjian Wen, University of Minnesota | Uncertainty quantification in atomistic simulations with dropout neural network potentials

4-4:20 p.m. – Jean Rabault, University of Oslo | Deep Reinforcement Learning reduces cylinder drag in a 2D flow simulation: a first step towards novel Active Flow Control methodology?

4:20-4:40 p.m. – Svetlana Tokareva, Los Alamos National Laboratory | Machine learning approach for the solution of the Riemann problem in fluid dynamics

4:40-5:10 p.m. – Updates | Dimitri Kusnezov, Department of Energy

5:15-7 p.m. – Poster Session

Wednesday, February 20, 2019

7-8:30 a.m. — Breakfast

8:30 a.m.-Noon – Session 3 | ML for Materials Microstructure

8:30-8:55 a.m. — Laurent Capolungo, Los Alamos National Laboratory | Machine Learning from physics-based spectral polycrystal plasticity models, Part I

8:55-9:20 a.m. – Ricardo Lebensohn, Los Alamos National Laboratory | Machine Learning from physics-based spectral polycrystal plasticity models, Part II

9:20-9:50 a.m. – Yong Han, Lawrence Livermore National Laboratory | Feedstock optimization using computer vision and machine learning techniques

9:50-10:20 a.m. – Hari Viswanathan, Los Alamos National Laboratory | Flow and Fracture in Microstructure Accelerated by Machine Learning

10:20-10:45 a.m. — Break

10:45-11:05 a.m. — Eric Mjolsness, University of California, Irvine | Operator algebra dynamics for spatially embedded labelled graphs

11:05-11:25 a.m. – Jason Albright, Los Alamos National Laboratory | Machine Learning-based Optimization Strategies for Artificial Viscosity, Part I

11:25-11:45 a.m. – Nathan Urban, Los Alamos National Laboratory | Machine Learning-based Optimization Strategies for Artificial Viscosity, Part II

11:45-Noon – Discussions

Noon-1:45 p.m. — Lunch 1:45-5:15 p.m. – Session 4 | ML for Surrogate Models

1:45-2:30 p.m. – Surya Kalidindi, Georgia Institute of Technology | Materials Knowledge Systems for Accelerated Materials Innovation

2:30-3 p.m. – Jaroslaw Knap, U.S. Army Research Laboratory | Accelerating Scale Bridging via Surrogate Modeling

3-3:30 p.m. – Break

3:30-4 p.m. – Alejandro Strachan, Purdue University | Data science tools to enhance atomistic simulations of materials and quantify uncertainties

4-4:20 p.m. – Daniel White, Lawrence Livermore National Laboratory | Two-Scale Topology Optimization Using Neural Network Surrogate Models

4:20-4:40 p.m. – Hesam Salehipour, Autodesk AI Laboratory/Univeristy of Toronto | Towards the next generation of sub-grid scale parameterization in ocean and climate models using deep learning

4:40-5 p.m. – Bryan Kaiser, Massachusetts Institute of Technology | Deep Learning for Turbulence Closures in the Deep Ocean

5-5:20 p.m. – Michael Glinsky, Sandia National Laboratories | The Mallat Scattering
Transform (MST) in high energy density plasmas: a new look at nonlinear,
multiscale physics in HED

5:20-5:30 p.m. – Discussions

Thursday, Februrary 21, 2019

7-8:30 a.m. — Breakfast

8:30 a.m.-Noon – Session 5 | ML for Fluid Flow & Turbulence

8:30-9:15 a.m. — Michael Chertkov, University of Arizona and Los Alamos National Laboratory | From Deep to Physics-Informed Learning of Turbulence: Diagnostics & Beyond

9:15-9:40 a.m. – Luc Peterson, Lawrence Livermore National Laboratory | Learning-based predictive models: a new approach to integrating large-scale simulations and experiments, Part I

9:40-10:05 a.m. – Brian Van Essen, Lawrence Livermore National Laboratory | Learning-based predictive models: a new approach to integrating large-scale simulations and experiments, Part II

10:05-10:25 a.m. — Break

10:25-10:55 a.m. – Jianxun Wang, University of Notre Dame | Physics-informed machine learning approach for augmenting turbulence models

10:55-11:15 a.m – Amir Barati Farimani, Carnegie Mellon University | Deep Learning Convective Transport

11:15-11:35 a.m – Brandon Morgan, Lawrence Livermore National Laboratory | Data-Augmented Modeling of Transition to Turbulence in a Rayleigh-Taylor Mixing Layer

11:35-11:55 a.m. – Nathan Miller, Sandia National Laboratories | Coordinate-Invariant Near-Wall Turbulence Modeling via Neural Networks

11:55-Noon – Discussions

Noon-1:45 p.m. — Lunch 1:45-5 p.m. – Session 6: ML for model reduction

1:45-2:30 p.m. – Karen Willcox, University of Texas at Austin | Projection-based Model Reduction: Formulations for Physics-based Machine Learning

2:30-3 p.m. – Paul Constantine, University of Colorado, Boulder | Exploratory Model Analysis: Or, how to determine whether machine learning is right for you

3-3:20 p.m. – Break

3:20-3:50 p.m. – Eurika Kaiser, University of Washington | Nonlinear system identification using sparsity-promoting techniques

3:50-4:10 p.m. – Brad Wolfe, Los Alamos National Laboratory | Measuring Asymmetry in ICF Images with a Neural Network

4:10-4:30 p.m. — Rohit Deshmukh, The Ohio State University | Sparse Representations of Complex Fluid Flows Over Spatially Global and Local Subspaces Through Machine Learning

4:30-4:50 p.m. – Devin Francom, Los Alamos National Laboratory | Parameter Tuning and Physics Model Comparison Using Statistical Learning

4:50-5 p.m. – Discussions & Conference Wrap-up

Poster Presentations

Bedros Afeyan | Improving the Performance of Plasma Kiunetic Simulations by Iteratively Learned Phase Space tiling: Variational Constrained Optimization Meet Machine Learning

Amir Barati Farimani | Deep Learning Convective Transport

Kyle Buchheit | Accelerating Computational Fluid Dynamics Using TensorFlow

Rohit Deshmukh | Sparse Representations of Complex Fluid Flows Over Spatially Global and Local Subspaces Through Machine Learning

Krishna Garikipati | Variational system identification of the partial differential equations governing pattern-forming physics: Inference under varying fidelity and noise

Michael Glinsky | The Mallat Scattering Transform (MST) in high energy density plasmas: a new look at nonlinear, multiscale physics in HED

Humberto Godinez | Assimilation of Dynamic Combined Finite Discrete Element Methods using the Ensemble Kalman Filter

Vitaliy Gyrya | Exploration of Machine Learning for Polynomial Root Finding

Chengkun Huang | Machine Learning for Turbulence in Supernovae

Benjamin Nebgen | Deep Neural Networks for Effective Hamiltonian Parameterization

Diane Oyen | Emulating Mesoscale Crack Propagation in Brittle Materials with a Probabilistic Markov Model

Anikesh Pal | Can Deep Learning Model Turbulence?

Gavin Portwood | Identification of localized instabilities in turbulent flows by convolutional neural networks

Jean Rabault | Deep Reinforcement Learning reduces cylinder drag in a 2D flow simulation: a first step towards novel Active Flow Control methodology?

Hesam Salehipour | Towards the next generation of sub-grid scale parameterization in ocean and climate models using deep learning

Robert Singleton Jr. | Matrix Inversion on the D-Wave

Mauricio Tano | Data-driven turbulence modeling for molten salt coolants

Nathan Urban | Machine Learning-based Optimization Strategies for Artificial Viscosity, Part II

Brad Wolfe | Measuring Asymmetry in ICF Images with a Neural Network

Jiaxin Zhang | Deep Learning Method to Accelerate Monte Carlo Simulation