Database Techniques to Improve Scientific Simulations
Indexing for online function approximation
Scientific simulations approximate real world physical phenomena using complex mathematical models. In most simulations, the mathematical model driving the simulation is computationally expensive to evaluate and must be repeatedly evaluated at several different parameters and settings. This makes running large-scale scientific simulations computationally expensive. A common method used by scientists to speed up simulations is to store model evaluation results at some parameter settings during the course of a simulation and reuse the stored results (instead of direct model evaluations) when similar settings are encountered in later stages of the simulation. Storing and later retrieving model evaluations in simulations can be modeled as a high dimensional indexing problem. Database techniques for improving scientific simulations focus on addressing the new challenges in the resulting indexing problem.
- 5.Panda B, Riedewald M, Gehrke J, Pope SB. High speed function approximation. In: Proceedings of the 7th IEEE International Conference on Data Mining; 2007.Google Scholar
- 6.Panda B, Riedewald M, Pope SB, Gehrke J, Chew LP. Indexing for function approximation. In: Proceedings of the 32nd International Conference on Very Large Data Bases; 2006.Google Scholar
- 9.Veljkovic I, Plassmann P, Haworth DC. A scientific on-line database for efficient function approximation. In: Proceedings of the International Conference on Computational Science and its Applications; 2003. p. 643–53.Google Scholar