SciML 2018 : DOE ASCR Scientific Machine Learning Workshop
Call For Papers
Call for Position Papers: Scientific Machine Learning
SciML2018: DOE ASCR Workshop on Scientific Machine Learning
North Bethesda, MD, United States, January 30-February 1, 2018
Deadline for submissions: January 5, 2018
This Call for Position Papers is issued for a workshop on Scientific Machine Learning as mentioned at the DOE Office of Science, Advanced Scientific Computing Research (ASCR) Advisory Committee meeting on September 26, 2017 (https://go.usa.gov/xnRyR).
Submissions will be reviewed by the Organizing Committee. Position papers will be selected based on their overall quality – see Submission Guidelines. Authors of selected submissions will be invited to participate in the workshop. Unique positions that are well-presented and emphasize potentially transformative research directions will be given preference. Authors are not expected to have a history of funding by the ASCR Research program. Authors are expected to actively participate in the workshop, stimulate constructive discussion by the participants, and contribute to an informative report. The workshop will be held on January 30 to February 1, 2018 in North Bethesda, Maryland.
• 5-Jan-2018: Position paper submission deadline
• 12-Jan-2018: Authors will be notified of selection for the anticipated workshop via the corresponding author on the abstract.
• 30-Jan-2018 to 1-Feb-2018: Anticipated workshop dates
Workshop URL: https://www.orau.gov/ScientificML2018
Submission URL: https://easychair.org/cfp/SciML2018
We are organizing a Workshop on Scientific Machine Learning on behalf of the Advanced Scientific Computing Research (ASCR) program in the Department of Energy (DOE) Office of Science and ASCR Program Manager Dr. Steven Lee. The ASCR program has a long tradition of providing high-performance computing (HPC) leadership for transforming science and energy research. ASCR accomplishes this by developing and maintaining world-class scientific computing and network facilities; advancing research in applied mathematics and computer science; and working in partnership with a broad range of researchers to solve increasingly complex challenges in computational and data sciences.
Recently, interest in machine learning-based approaches for science and engineering has soared. This growing interest is due to the combined use of efficient analysis algorithms, massive amounts of data available from scientific instruments and other sources, advances in high-performance computing, and the successes reported by industry, academia, and research communities. A conventional notion of machine learning involves the training of an algorithm to automatically find patterns, signals, or structure that may be hidden within massive data sets, and too well-hidden to be programmed for explicitly. The predictive capability of the algorithm is a learned skill. We seek to improve upon and harness the predictive power of machine learning to maximize its impact on DOE-mission and science/engineering applications.
The purpose of the workshop will be to define priority research directions for applied mathematics in scientific machine learning. In particular, the workshop will define the challenges and opportunities for increasing the rigor, robustness, and reliability of machine learning for DOE science and engineering applications. In terms of examples, such challenges might include rigorous analysis methods for developing and testing machine learning methods, understanding machine learning method approximation power, bounding data and compute complexity of machine learning approaches, etc. These challenges will be discussed in the context of existing methods with demonstrated use in scientific applications. The workshop will not focus on the development of new machine learning methods.
We encourage participation from a wide range of institutions including, but not limited to, universities, industry, and DOE National Laboratories. The workshop will feature a variety of plenary talks and multiple break-out sessions, with every invitee expected to participate actively in discussion of potential research directions. We anticipate that participants will produce a report that will define basic needs and opportunities in applied mathematics research for scientific machine learning.
We invite community input in the form of two-page position papers that identify and discuss key applied mathematics challenges posed for scientific machine learning. In addition to providing an avenue for identifying workshop participants, the position papers will be used to shape the workshop agenda, identify panelists, and contribute to the workshop report. Position papers should not describe the authors’ current or planned research, nor should they recommend solutions or narrowly focused research topics. Rather, they should aim to improve the community’s shared understanding of the scientific machine learning problem space and help to stimulate discussion.
Position papers should describe relevant existing, or opportunities for, applied mathematics research that has a clear potential to enhance the mathematical understanding of machine learning, particularly in terms of the rigor and robustness of scientific machine learning. The papers should clearly address the following points:
• Key Challenges: What challenges related to mathematical aspects of scientific machine learning is the paper intended to address?
• New Research Directions: What is the potential for new mathematical research to address the challenge?
• State of the Art: Set the new ideas in context by describing the state of the art relating to the proposed research direction.
This description should be followed by an assessment of potential research directions based on the following dimensions:
• Maturity: Are there existing mathematical methods or research directions that address the challenge(s) and that show promise for scientific machine learning? What are the indicators that a given method or approach will address the identified challenges? If there are not existing methods or research approaches to meeting the challenge, can you suggest ways to gain new insight into the problem space?
• Uniqueness: Is the identified challenge unique to scientific applications of machine learning? What makes it so?
• Novelty: Is the challenge being addressed by other research programs? By the private sector? Why should this challenge be of interest to the ASCR Applied Mathematics program?
Each position paper must be no more than two pages, including figures and references. The paper may include any number of authors, but must provide contact information for a single author, who could represent the position paper at the workshop. There is no limit to the number of position papers that an individual or group can submit. Authors are strongly encouraged to follow the structure outlined above. Submit position papers in PDF format at https://easychair.org/cfp/SciML2018