This page is not currently being maintained and is for archival information only.
Please see our most recent work at our Information Technology Laboratory homepage, where ITL’s Purpose IS Cultivating Trust in IT and Metrology
Modern scientific research has become more and more dependent on mathematical, statistical, and computational tools for enabling discovery. The Enabling Scientific Discovery Program promotes the use of these tools to dramatically advance our ability to predict the behavior of a broad range of complex scientific and engineering systems and enhance our ability to explore fundamental scientific processes. This Program focuses on interdisciplinary scientific projects that involve novel computational statistics and the development of simulation methods and software. These efforts will have a foundational impact on scientific discovery throughout U.S. industry, government, and academia.
The development of new methods for simulating and modeling complex phenomenon in science has become the area yielding the greatest current breakthroughs. From the modeling and simulation of material science phenomenon (materials with memory for healthcare applications) to the modeling and simulation of quantum dynamics for use in computation and communication, these fields are dense with difficult problems. The problems are difficult because the underlying sub-fields have progressed so far in such a short period of time: e.g. physicists now want to predict phenomenon on smaller and smaller time and space scales and do so with greater speed and accuracy, chemists now must employ spectroscopy to understand subtle differences in pharmaceutical isomers, a process that requires noise modeling and simulation, extremely difficult quantum phenomenon must be modeled and simulated accurately to predict our next generation of computer processes and the associated computer commerce security issues. This is clearly a basket of difficult problems that are, quite simply, not solved anywhere else. The new ideas being employed by ESD contributing NIST staff members is a combination of novel computational and visualization techniques, numerical and algorithmic developments, and a unique collaboration with field-specific scientists.
The need for a computational statistics research effort within ITL is a direct consequence of NIST's role as measurement laboratory. Simply put, there is no way to perform measurement science without a two-pronged statistics research effort: the study of novel forms of applications of computational statistics, and the development of novel statistical techniques. Both aspects are necessary for any measurement science to be accomplished. The primary long term goal of the Computational Statistics project of ESD is the construction and maintenance of a portfolio of statistics-enhanced scientific work. Everything from measurement of uncertainty, the study of sensitivity as it pertains to calibration, the use of multi-dimensional scaling and of course probability and distributions (for example maximum likelihood density estimation) must all be application practiced and generally advanced on a regular basis.