Xiaoming Huo, Program Director for Statistics, Computational and Data-enabled Science & Engineering at NSF, writes:

CDS&E-MSS (Computational and Data-Enabled Science and Engineering in Mathematical and Statistical Sciences) is a relatively new funding opportunity that is managed by the Division of Mathematical Sciences (DMS) of the National Science Foundation (NSF). Its updated information can be found at the program web site (http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=504687). The next submission window will be November 25–December 9, 2014.

The CDS&E-MSS program supports research that confronts the host of mathematical and statistical challenges presented to the scientific and engineering communities by the ever-expanding role of computational modeling and simulation on the one hand, and the explosion in production of digital and observational data on the other. The goal of the program is to promote the creation and development of the next generation of mathematical and statistical theories and methodologies that will be essential for addressing such issues. To this end, the program supports fundamental research in mathematics and statistics whose primary emphasis is on meeting these computational and data-related challenges.

CDS&E-MSS can be viewed as a complement to the Big Data endeavor of NSF (http://www.nsf.gov/news/news_summ.jsp?cntn_id=123607) as well as a partner of the Data-to-Knowledge-to-Action drive by the Office of Science and Technology Policy (White House).

CDS&E-MSS is highly integrated and collaborative with CDS&E, which is a larger effort in NSF and involves many other NSF directorates and divisions. We refer you to http://www.nsf.gov/mps/cds-e/ and http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=504813 for further information on CDS&E. Note that CDS&E-MSS and CDS&E are not the same; if proposed work emphasizes mathematical or statistical development, CDS&E-MSS may be a fit. If the proposed work is more driven by particular scientific and/or engineering applications, the CDS&E program (which covers wider areas) may be more suitable. Also please note that CDS&E and CDS&E-MSS have different submission windows and different solicitation numbers. Potential investigators should contact the program directors to assess the suitability of their projects to a particular program.

You may have seen information on this program via a Dear Colleague letter at www.nsf.gov/pubs/2012/nsf12018/nsf12018.jsp, and an October 2012 article by Jia Li in Amstat News.

CDS&E-MSS was launched in 2011. The first two rounds of awards were made in the middle of 2012 and 2013, respectively. To know more about CDS&E-MSS supported projects, one can visit the NSF/DMS web site, and locate the link to CDS&E-MSS. After arriving at the CDS&E-MSS web site, at the bottom of the web site, there is a hyperlink named “What Has Been Funded…” which takes you to the NSF Award Search web site, where all active projects funded by the CDS&E-MSS program are listed. An alternative means is to go to the NSF Award Search web site and then use the CDS&E program element code 8069 in search.

The awards from CDS&E-MSS cover a wide range of topics: e.g., stochastic partial differential equations, Lie groups and representation theory, manifold learning, sparse optimization, data assimilation, partially-observed Markov processes, and high dimensional learning. Many emerging methodologies have been proposed to be developed: e.g., efficient parallel iterative Monte Carlo methods, accelerated Monte Carlo schemes, solving large-scale eigen-related problems, and measurement model specification search. Some projects are dealing with newly emerged datasets: e.g., algebraic, geometric, and computational tools for data cloud and data array; LiDAR point cloud data; and data with network structure. A wide range of applications can be found in the current awards, including tumor microenvironment, genetic association, brain connectivity, coastal ocean modeling, and subsurface imaging. More information can be found online.

Many CDS&E-MSS awards support interdisciplinary research. In fact, a large proportion of existing awards are jointly supported by several divisions within NSF.

Principal investigators are advised not to submit to CDS&E-MSS a proposal for research that could be supported by another DMS program. For CDS&E-MSS, besides addressing the data-enabling component (which seems to be a strength of many statistically oriented projects), it is equally important to argue convincingly on its application(s) in science and engineering. A project that appears to fit into other traditional programs will have low funding priority in the CDS&E-MSS program.

The statistics community ought to play a significant role in the CDS&E-MSS. Note the letter D (in CDSE) stands for “data-enabled.” Because statistics is a significant component of data science, statisticians are well positioned to tackle the problems that fit into CDS&E-MSS. Proposers are more likely to succeed if they demonstrate their efforts to collaborate with other scientists, who can serve as domain experts in the associated science and engineering fields.

In summary, if you propose a research project for development of the next generation of statistical theories and methodologies that will address computational and data-related challenges, with solid collaboration and important scientific and engineering application, CDS&E-MSS is your potential sponsor.