Statistics Visualization Modeling Projects News

Unprecedented advances in computation and storage have enormously improved abilities to model complex processes in science, engineering and the social sciences. In parallel, data sets and observational sets have grown in size and complexity beyond what could have been imagined a few decades ago. Gaining knowledge and insight from these efforts requires detailed rigorous comparison of models and data, and the ever increasing sophistication of the models and size and detail of the heterogenous data sets demands commensurate advances in the statistical analysis of data. The MADAI collaboration (Models and Data Analysis Initiative) is devoted to developing statistical analysis and visualization tools for the purpose of comparing complex models and simulations to large scale heterogenous data sets. Since many of the challenges cross a multitude of disciplines, MADAI's analysis infrastructure will be designed to simultaneously address several pressing scientific challenges, and once finished, should be extensible to innumerable other problems. The initial modeling challenges come from simulating nuclear collisions, galaxy formation, meso-scale atmospheric dynamics and bio-chemical evolution. Analyses will rigorously determine fundamental parameters, such as the viscosity of the quark gluon plasma, or the percentage of dark matter in the Milky Way. The most accurate and efficient methods for modeling complicated processes such as the dynamics of the atmospheric boundary level will be determined by rigorously comparing models to data. Furthermore, the most important links between model parameters or assumptions and key observables will be identified by exploiting statistical and visualization tools. The MADAI collaboration is funded by the Cyber-Enabled Discovery and Innovation (CDI) program of the National Science foundation for a four-year period beginning in the Fall of 2009. The multidisciplinary collaboration includes physicists, cosmologists, atmospheric scientists, statisticians and visualization experts from Michigan State University, Duke University and the University of North Carolina.

Home Members Internal