Workshop Program

AI4S 2022 will be held with the hybrid virtual/in-person mode at 1:30-5pm on Monday, November 14 at C144-145. The schedule is provided below. All times are in CST (the timezone of Dallas, TX, USA).

1:30PM	Welcome and Intro
1:30-2:05PM	AI4S Featured Speaker: Prof. Geoffrey Fox (University of Virginia) Session chair: Gokcen Kestor
2:05-2:20PM	Paper: A Case Study on Coupling OpenFOAM with Different Machine Learning Frameworks (Fabian Orland, Kim Sebastian Brose, Julian Bissantz, Federica Ferraro, Christian Terboven, and Christian Hasse.) Session chair: Dong Li
2:20-2:35PM	Paper: Automated Continual Learning of Defect Identification in Coherent Diffraction Imaging (Orcun Yildiz, Henry Chan, Krishnan Raghavan, William Judge, Mathew J. Cherukara, Prasanna Balaprakash, Subramanian Sankaranarayanan, and Tom Peterka.) Session chair: Dong Li
2:35-2:50PM	Paper: Pattern-Based Autotuning of OpenMP Loops Using Graph Neural Networks (Akash Dutta, Jordi Alcaraz, Ali Tehrani Jamsaz, Anna Sikora, Eduardo Cesar, and Ali Jannesari.) Session chair: Dong Li
2:50-3PM	Paper: Determining HEDP Foams' Quality with Multi-View Deep Learning Classification (Nadav Schneider, Matan Rusanovsky, Raz Gvishi, and Gal Oren.) Session chair: Dong Li
3-3:30PM	Break
3:30-3:45PM	Paper: Ensuring AI For Science Is Science: Making Randomness Portable (Hana Ahmed, Roselyne Tchoua, and Jay Lofstead) Session chair: Wenqian Dong
3:45-4PM	Paper: Practical Federated Learning Infrastructure for Privacy-Preserving Scientific Computing (Lesi Wang and Dongfang Zhao) Session chair: Wenqian Dong
4-4:10PM	Paper: Scalable Integration of Computational Physics Simulations with Machine Learning (Mathew Boyer, Wesley Brewer, Dylan Jude, and Ian Dettwiller) Session chair: Wenqian Dong
4:10-4:20PM	Paper: PhySRNet: Physics Informed Super-Resolution Network for Application in Computational Solid Mechanics (Rajat Arora Arora) Session chair: Wenqian Dong
4:20-5PM	Panel (Feiyi Wang (ORNL), Pawan Balaji (Meta), Laurent White (AMD), and Debbie Bard (LBNL)) Moderator: Jay Lofstead

Call for Papers

The purpose of this workshop is to bring together computer scientists and domain scientists from academia, government, and industry to share recent advances in the use of AI/ML to various scientific applications, introduce new scientific application problems to the broader community, and stimulate tools and infrastructures to support the application of AI/ML in scientific applications. The workshop will be organized as a series of plenary talks based on peer-reviewed paper submissions accompanied by keynotes from distinguished researchers in the area and a panel discussion. We encourage participation and submissions from universities, industry, and DOE National Laboratories.

Artificial intelligence (AI)/machine learning (ML) is a game-changing technology that has shown tremendous advantages and improvements in algorithms, implementation, and applications. We have seen many successful stories of applying AI/ML to scientific applications, such as predicting extreme weather events, identifying exoplanets in trillions of sky pixels, and accelerating numerical solvers in fluid simulation. However, there are a number of problems remaining to be studied to enhance the usability of AI/ML to scientific applications. For example, how to systematically and automatically apply AI/ML to scientific applications? How to incorporate domain knowledge (e.g., conservation laws, invariants, causality and symmetries) into AI/ML models? How to make the models interpretable and robust for HPC? How to make AI/ML more approachable to the HPC community? Addressing the above problems will bridge the gap between AI/ML and scientific applications and enable wider employment of AI/ML in HPC.

Topics will include but will not be limited to:

Innovative AI/ML models to analyze, accelerate, or improve performance of scientific applications in terms of execution time and simulation accuracy;

Innovative methods to incorporate complex constraints imposed by physical principles to scientific applications;

Innovative methods to completely or partially replace first-order computation with efficient AI/ML models;

Tools and infrastructure to improve the usability of AI/ML to scientific applications;

Performance characterization and study on the possibility of using AI/ML to specific scientific applications;

Workflow of applying AI/ML to scientific applications;

Innovative methods to make AI models interpretable and robust for scientific applications.

Submission

Authors are invited to submit manuscripts in English structured as technical papers up to 6 pages, letter size (8.5in x 11in) and including figures, tables, and references. Submissions not conforming to these guidelines may be returned without review. Your paper should be formatted using IEEE conference format which can be found from here

All manuscripts will be peer-reviewed and judged on correctness, originality, technical strength, and significance, quality of presentation, and interest and relevance to the workshop attendees. Submitted papers must represent original unpublished research that is not currently under review for any other conference or journal. Papers not following these guidelines will be rejected without review and further action may be taken, including (but not limited to) notifications sent to the heads of the institutions of the authors and sponsors of the conference. Submissions received after the due date, exceeding length limit, or not appropriately structured may also not be considered. At least one author of an accepted paper must register for and attend the workshop. Authors may contact the workshop organizers for more information.

Papers should be submitted electronically at: https://submissions.supercomputing.org, SC22 Workshop: AI4S'22: Workshop on Artificial Intelligence and Machine Learning for Scientific Applications".

The final papers are planned to be published through IEEE. Published proceedings will be included in the IEEE Xplore digital library.

Important Dates