Overview

• The BIE provides general framework for Bayesian inference, tailored to astronomical and earth-science survey data. Its primary uses are
  • Modeling: parameter determination over terabyte-scale data sets
  • Data mining: general non-parametric characterization
• It can incorporate multiple survey datasets simultaneously
• Inference by Bayesian Monte Carlo Markov Chain (MCMC).
• It uses directed sampling and a new approach--hierarchical empirical priors--that efficiently explores and limits high-dimensional parameter space
• Currently being used to infer Galactic structure from star counts but can be used for any scientific dataset to compare with predicted distributions. The system is ideal for mapped data (spatial/temporal attributes) for example:
  • Galactic ``spectral'' data cubes from astronomical radio surveys
  • Multi-wavelength observations of nearby external galaxies
  • Earth Observation System (EOS) data sets and combination with existing geographical and geophysical surveys