Bulletin of the American Physical Society
APS March Meeting 2023
Volume 68, Number 3
Las Vegas, Nevada (March 5-10)
Virtual (March 20-22); Time Zone: Pacific Time
Session Y53: Open Science and Data SetsFocus
|
Hide Abstracts |
Sponsoring Units: GDS FIAP Chair: Savannah Thais, Columbia University Room: Room 307 |
Friday, March 10, 2023 8:00AM - 8:36AM |
Y53.00001: Enabling AI for Open Science on Supercomputers at NERSC Invited Speaker: Steven Farrell AI is a transformative force for computational science that can enhance and accelerate scientific discovery. As these methodologies and applications mature and grow more complex, they increasingly require high performance computing (HPC) resources for developing and deploying models. NERSC is the mission HPC facility for the DOE Office of Science and provides advanced supercomputing resources for open science across many domains including materials sciences, biosciences, earth sciences, and physics. This presentation will describe our vision at NERSC for enabling and supporting cutting-edge scientific AI through advanced HPC system deployments, research engagements, benchmark challenges and datasets, and outreach. I will discuss our efforts to develop and support benchmarks in MLCommons for HPC and Science that are driving innovation in performance and scientific discovery, as well as efforts to provide a productive platform and ecosystem for user workloads and benchmarks. Finally, I will describe our various outreach activities such as surveys which inform us of the AI compute needs of our community and training events which seek to educate these communities in the methods and tools to effectively utilize HPC and AI for today's greatest scientific challenges. |
Friday, March 10, 2023 8:36AM - 8:48AM |
Y53.00002: Facilitating Open Science Practices at User Facility Lyudmila Balakireva, Fedor F Balakirev We will describe a data management framework deployed at the National High Magnetic Field Laboratory (NHMFL/MagLab) in collaboration with the Los Alamos National Laboratory (LANL) Research Library Prototyping team. The framework is aimed at helping scientists efficiently manage experimental data produced by their MagLab projects by automatically incorporating the data into the permanent record at the Open Science Framework platform developed by the Center for Open Science. We will describe elements of the data framework such as data formats, metadata schema, repository to use, naming conventions, template to organize the data for the experiment, and automated data pipeline from instruments to the FAIR repository, reducing the entry barrier for the acceptance and proliferation of the open science practices. Work at NHMFL-LANL was performed under the auspices of the NSF, DoE, and State of Florida. |
Friday, March 10, 2023 8:48AM - 9:00AM |
Y53.00003: Application of causality-first functional decomposition trees to data science and materials informatics Hiori Kino The promotion of open science, in which data generated from experimental and theoretical calculations are released in accordance with the FAIR principle to reuse the data, is a worldwide trend. It is desirable to use metadata with vocabulary of common or interconvertible conceptual structures to publish data. Furthermore, the conceptual structure of many data schemas is created to form a hierarchical structure, but since there is no single conceptual structure, it is more convenient to create the conceptual structure based on a method that allows terms to be easily added or modified. |
Friday, March 10, 2023 9:00AM - 9:12AM Author not Attending |
Y53.00004: A Dynamics Data Set for Spin Ice Kyle G Sherman, Michael J Lawler Spin ice is a magnetic system whose constituents fluctuate collectively to produce emergent magnetic monopoles at low temperatures. These monopoles affect an equilibration which can take days to complete, despite microsecond spin-flip processes. In the interest of solving this mysterious long-time dynamical behavior in spin ice materials, we have generated a benchmark stochastic time series data set for both pyrochlore spin ice Dy2Ti2O7 and artificial checkerboard spin ice. We present the results of our analysis which implements both traditional and data-driven methods. We further show that machine learning methods are capable of learning from these data sets by training a deterministic convolutional neural network that can statistically reproduce the stochastic data. We hope to provide a path for the study of frustrated and topological magnetic systems open to the data science community. |
Friday, March 10, 2023 9:12AM - 9:24AM |
Y53.00005: Novel Representations and Quantitative Structure Property Relationships for Polymers Using Machine Learning Javad Tamnanloo, Everen J Wegner, Abraham Joy, Mesfin Tsige Current developments indicate Machine Learning (ML) can be utilized to identify, design, and optimize polymeric structures for desired properties, vastly reducing the costly and time-consuming experimentation required. However, Polymer Informatics faces the significant challenge of accurately and effectively representing large and complex polymeric structures in a computer-comprehensible manner. To create computationally efficient models, we developed novel representations of polymers that simplify yet retain the molecular structure. From each representation, over 1,800 quantitative structural chemical descriptors were generated, and a novel approach was used to better emulate the structure of long polymers and counteract the compressional effects of the representation used with less computational cost. Pruned subsets of these descriptors were utilized to train over 25 different types of ML Models with Bayesian Hyperparameter Optimization and were evaluated by predicting Glass Transition Temperature (Tg) for 564 polymers and were shown to achieve r-squared values up to 0.74. These promising preliminary results indicate that given the proper dataset, this process could be utilized to create ML models for any other desired property(s). |
Friday, March 10, 2023 9:24AM - 9:36AM |
Y53.00006: Application of GARCH Models Generalized with Machine Learning and Quantum Statistical Physics Methodology in Exotic Financial Data and Beyond Haidong Yan With the recent development of econophysics, information theory, and correlated system mechanics, many random and stochastic processes in complex systems can be much better described, such as the financial market, the infectious disease dynamics, the social network construction, and others. The generalized autoregressive conditional heteroskedasticity (GARCH) model is a well-established method in analyzing and predicting the volatility of a fluctuating system. However, within the traditional GARCH scenario, the parameter fitting and optimization is mostly limited to the linear space. |
Friday, March 10, 2023 9:36AM - 10:12AM |
Y53.00007: NexusLIMS: A Laboratory Information Management System for Shared-Use Electron Microscopy Facilities Invited Speaker: June W Lau Presently, the management and the processing of research data produced by scientific instrumentation is a significant challenge. Electron microscopy researchers work with different make and model of microscopes and detectors, and with these, data produced in various (sometimes proprietary) formats. Users often implement ad hoc strategies to curate and disseminate this data. Often, data and the context under which they were acquired is irrecoverable beyond the few published images in journal articles. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700