Statistics – Computation
Scientific paper
Jan 2010
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2010aas...21523004s&link_type=abstract
American Astronomical Society, AAS Meeting #215, #230.04; Bulletin of the American Astronomical Society, Vol. 42, p.579
Statistics
Computation
Scientific paper
The emerging large datasets are posing major challenges for their subsequent statistical analyses. One needs reinvent optimal statistical algorithms, where the cost of computing is taken into account. Moving large amounts of data is becoming increasingly untenable, thus our computations must be performed close to the data. Existing computer architectures are CPU-heavy, while the first passes of most data analyses require an extreme I/O bandwidth. Novel computational algorithms, optimized for extreme datasets, and the new, data-intensive architectures must be invented. The outputs of large numerical simulations increasingly resemble the "observable” universe, with data volumes are approaching if not exceeding observational data. Persistent "laboratories” of numerical experiments will soon be publicly available, and will change the way we approach the comparisons of observational data to first principle simulations.
No associations
LandOfFree
Data-Intensive Statistical Computations in Astronomy does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Data-Intensive Statistical Computations in Astronomy, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Data-Intensive Statistical Computations in Astronomy will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-961017