Structure Discovery from Massive Spatial Data Sets Using Intelligent Simulation Tools

  • Authors:

  • Feng Zhao;Chris Bailey-Kellogg;Xingang Huang;Iván Ordóñez

  • Affiliations:

  • Palo Alto Research Center (PARC), Palo Alto, California, USA;Department of Computer Sciences, Purdue University, West Lafayette, Indiana, USA;The Ohio State University, Columbus, Ohio, USA;Bios Group, Santa Fe, New Mexico, USA

  • Venue:
  • Computational Discovery of Scientific Knowledge
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Extracting structures as communicable knowledge is a central problem in spatio-temporal data analysis. Spatial Aggregation is an effective way for discovering structures. To address the computational challenges posed by applications such as weather data analysis or engineering optimization, Spatial Aggregation recursively aggregates local data into higher-level descriptions, exploiting the fact that these physical phenomena can be described as spatio-temporally coherent "objects" due to continuity and locality in the underlying physics. This paper uses several problem domains -- weather data interpretation, distributed control optimization, and spatio-temporal diffusion-reaction pattern analysis -- to demonstrate that intelligent simulation tools built upon the principles of Spatial Aggregation are indispensable for scientific discovery and engineering analysis.