Events

Automatic interpretation and control of X-ray diffraction experiments

The emergence of high-throughput quantum chemical calculations has accelerated the rate at which we can predict new materials for various applications (batteries, solar cells, catalysts, etc.), but the successful synthesis of these materials has often become the slow step in materials design. 

Autonomous laboratories hold the potential to systematically explore various synthesis routes to new materials, alleviating the painstaking manual trial-and-error approach. However, for an autonomous laboratory to work for inorganic synthesis, we need a method for assessing the success of a given synthesis effort without any human intervention. 

Powder X-ray diffraction (XRD) is the workhorse technique for determining the outcome of materials synthesis. In this talk, I will show how convolutional neural networks can be trained to automatically interpret XRD patterns and identify the phases present in realistic mixtures of crystalline phases. I will also discuss how these same models can be leveraged to adaptively control the XRD experiment itself to improve the quality of predictions and enable the detection of short-lived intermediates that form during synthesis.

About Chris Bartel
Chris Bartel joined the Department of Chemical Engineering and Materials Science (CEMS) at the University of Minnesota as an Assistant Professor in August 2022. His research group is using quantum chemical calculations and machine learning to accelerate the discovery and design of solid-state materials for sustainable energy technologies. Prior to his current appointment, he was a postdoc in Materials Science & Engineering at UC Berkeley and Berkeley Lab, working with Prof. Gerd Ceder. He earned his PhD in Chemical Engineering at the University of Colorado Boulder under the supervision of Prof. Al Weimer and Prof. Charles Musgrave. He was the recipient of an NSF Graduate Research Fellowship and the Max S. Peters Outstanding Graduate Award at the University of Colorado.
 

Physics-Informed Deep Learning: Blending data and physics for fast predictions

Distinguished Speaker: George Em Karniadakis, the Charles Pitts Robinson and John Palmer Barstow Professor of Applied Mathematics and Engineering at Brown University, Research Scientist at Massachusetts Institute of Technology, and Director of the Physics-Informed Learning Machines for Multiscale and Multiphysics Problems (PhILMs) at the Pacific Northwest National Laboratory

In this lecture, we will review physics-informed neural networks and summarize available extensions for applications in computational mechanics and beyond. We will also introduce new NNs that learn functionals and nonlinear operators from functions and corresponding responses for system identification. The universal approximation theorem of operators is suggestive of the potential of NNs in learning from scattered data any continuous operator or complex system. We first generalize the theorem to deep neural networks, and subsequently we apply it to design a new composite NN with small generalization error, the deep operator network (DeepONet), consisting of a NN for encoding the discrete input function space (branch net) and another NN for encoding the domain of the output functions (trunk net). We demonstrate that DeepONet can learn various explicit operators, e.g., integrals, Laplace transforms and fractional Laplacians, as well as implicit operators that represent deterministic and stochastic differential equations. More generally, DeepOnet can learn multiscale operators spanning across many scales and trained by diverse sources of data simultaneously.

About the speaker
George Karniadakis is from Crete. He is a member of the National Academy of Engineering and a Vannvar Bush Faculty Fellow. He received his S.M. and Ph.D. from Massachusetts Institute of Technology (1984/87). He was appointed Lecturer in the Department of Mechanical Engineering at MIT and subsequently he joined the Center for Turbulence Research at Stanford/Nasa Ames. He joined Princeton University as Assistant Professor in the Department of Mechanical and Aerospace Engineering and as Associate Faculty in the Program of Applied and Computational Mathematics. He was a Visiting Professor at Caltech in 1993 in the Aeronautics Department and joined Brown University as Associate Professor of Applied Mathematics in the Center for Fluid Mechanics in 1994. After becoming a full professor in 1996, he continued to be a Visiting Professor and Senior Lecturer of Ocean/Mechanical Engineering at MIT. He is an AAAS Fellow (2018-), Fellow of the Society for Industrial and Applied Mathematics (SIAM, 2010-), Fellow of the American Physical Society (APS, 2004-), Fellow of the American Society of Mechanical Engineers (ASME, 2003-) and Associate Fellow of the American Institute of Aeronautics and Astronautics (AIAA, 2006-). He received the SIAM/ACM Prize on Computational Science & Engineering (2021), the Alexander von Humboldt award in 2017, the SIAM Ralf E Kleinman award (2015), the J. Tinsley Oden Medal (2013), and the CFD award (2007) by the US Association in Computational Mechanics. His h-index is 123 and he has been cited over 70,000 times.

AI for Digital Pathology

Pathology plays a critical role in the diagnosis of disease and is the basis of all tissue-based biomarkers. In recent years, the advancements of AI in digital pathology have shown immense potential in improving patient outcomes through accurate diagnosis and enabling drug development. This talk introduces AI for digital pathology and covers pathology domain-specific machine learning challenges. We briefly spotlight Multiple Instance Learning (MIL) for pathology and a recent work by PathAI to improve its interpretability. We also explore Test Time Adaptation, a form of domain generalization, in the context of pathology. ML methods like these solve core problems in the field and enable PathAI’s products to succeed in clinical applications.

View the presentation slides

About the speakers and PathAI
Dinkar Juyal is a Machine Learning Engineer at PathAI. For more than 3 years, his work at PathAI has centered around topics such as weak supervision, interpretability, robustness of ML models, as well as building ML products for pathology. He received his masters's degree in Operations Research from University of California, Berkeley in 2019. His work has been published in multiple ML and medical conferences and journals, including NeurIPS, ICLR and Journal of Hepatology.

Andrew Walker joined PathAI as a Machine Learning Engineer in August. He graduated with a master's in computer science from the UMN Department of Computer Science and Engineering in May 2022. He completed his thesis, titled “Adaptive Domain Generalization for Digital Pathology Images” in Ju Sun’s GLOVEX group, in partnership with PathAI. 

PathAI is a leading provider of AI-powered research tools and services for pathology. PathAI's platform promises substantial improvements to the accuracy of diagnosis and the efficacy of treatment of diseases like cancer, leveraging modern approaches in machine and deep learning. Based in Boston, PathAI works with leading life sciences companies and researchers to advance precision medicine. To learn more, visit pathai.com.

Wonjun Lee (University of Minnesota, Twin Cities) presents a method to efficiently compute Wasserstein gradient flows. Their approach is based on a generalization of the back-and-forth method (BFM) introduced by Jacobs and Leger to solve optimal transport problems.

View the IMA event page for more information

The program will include faculty and student presentations, as well as a poster session covering wide-ranging topics in bioinformatics and computational biology.

More information will be available in the future.

Have you ever wondered:
How does Netflix know what movies you like?
How does your phone know what to suggest in your next text?
How do professional sports teams draft players with statistics?

Join us at camp to learn about the answers to questions like these and more!

• Use the Python programming language to analyze financial and social network data that can be found around us on a daily basis
• Apply machine learning algorithms to projects of your choice

This is a unique opportunity to connect with other high school students who share the same interests, and learn cool programming and math techniques that make a difference in the world.

There will be more information added in the future. Check the IMA event page for more information.

Federated machine learning (FL) is gaining a lot of traction across research communities and industries. FL allows machine learning (ML) model training without sharing data across different parties, thus natively supporting data privacy. However, designing and executing FL jobs is not an easy task today. Flame is an open-source project that aims to ease the composition of FL jobs and the management of their lifecycle across different environments. Towards those ends, Flame is architected to be open and extensible from its inception. This talk will present an overview of the project and a demo on how the Flame system works in a Kubernetes environment.

About Myungjin Lee
Myungjin Lee is a Senior Researcher at Cisco's Emerging Technologies and Incubation (ET&I). He leads research on systems for edge computing. His current focus is on federated learning and its use cases at the edge. He is passionate about building software for distributed systems and computer networks.

Prior to Cisco, he worked at Salesforce as a software engineer, where he led a secure cross-datacenter communication project. He was also an Assistant Professor at the University of Edinburgh, UK, where he led research activities around systems and networks including datacenter networks, network telemetry, SDN, etc. 

There will be more information added in the future. Check the IMA event page for more information.

Last month, IPUMS PMA released the harmonized version of Performance Monitoring for Action's redesigned core family planning survey, which includes a longitudinal panel of childbearing women for analyzing contraceptive and fertility dynamics over time. Join to learn how to download a dataset with panel data already linked and how to analyze these new data.

IPUMS provides census and survey data from around the world integrated across time and space. IPUMS integration and documentation makes it easy to study change, conduct comparative research, merge information across data types, and analyze individuals within family and community contexts. Data and services available free of charge.