In order to enhance professional relationships between academia and industry, the Statistics Department at the University of Georgia holds an annual "UGA Statistics Club Industry Day." This event helps the advancement of cutting edge research in statistics by bringing together a senior researcher from industry and students and faculty from academia. Speakers deliver two presentations: a technical afternoon research lecture (as a part of our regular seminar series) and a less technical after-lunch talk followed by a question-and-answer session with the students.


This Year's Abstract & Speaker:

Kary Myers is a scientist in the Statistical Sciences group at Los Alamos National Laboratory (www.stat.lanl.gov) and the deputy director for data science of the Information Science and Technology Institute (isti.lanl.gov). With an AT&T Labs Fellowship, she earned her PhD from Carnegie Mellon's Statistics Department and her MS from their Machine Learning Department. At Los Alamos she's been involved with projects examining electromagnetic measurements, large scale computer simulations, and chemical spectra from the Mars Science Laboratory Curiosity Rover. She serves as an associate editor for the Annals of Applied Statistics and the Journal of Quantitative Analysis in Sports, and she created and organizes CoDA, the Conference on Data Analysis (cnls.lanl.gov/coda).

Partitioning a Large Simulation as It Runs 

As computer simulations continue to grow in size and complexity, they present a particularly challenging class of big data problems. Many application areas are moving toward exascale computing systems, systems that perform a billion billion FLOPS (FLoating-point Operations Per Second). Simulations at this scale can generate output that exceeds both the storage capacity and the bandwidth available for transfer to storage, making post-processing and analysis challenging. One approach is to embed some analyses in the simulation while the simulation is running --- a strategy often called in situ analysis --- to reduce the need for transfer to storage. Another strategy is to save only a reduced set of time steps rather than the full simulation. Typically the selected time steps are evenly spaced, where the spacing can be defined by the budget for storage and transfer. Our work combines both of these ideas to introduce an online in situ method for identifying a reduced set of time steps of the simulation to save. Our approach significantly reduces the data transfer and storage requirements, and it provides improved fidelity to the simulation to facilitate post-processing and reconstruction. We illustrate the method using a computer simulation that supported NASA's 2009 Lunar Crater Observation and Sensing Satellite mission.

Statistics in the National Laboratories (Lunch Talk)

Dr. Myers will discuss what it's like to be a statistician at Los Alamos National Laboratory and at some of the other national labs run by the Department of Energy. She will minimize her prepared content in order to allow plenty of time for questions and discussion.