Project Elementary


Elementary 0.3 (released May 2013)

Added support for OpenBUGS as a model specification language [pre-alpha release]

  • Support OpenBUGS as a model specification language.

Elementary 0.2 (released Nov 2012)

Add-in support for Accumulo and HBase.

  • Support Accumulo and HBase as Storage Backends.

Elementary 0.1 (released Aug 2012)

This is the initial release of Elementary. Main features include:

  • Scalable single-machine Gibbs sampling: Takes advantage of hybrid storage manager to scale up problems that are larger than main memory.
  • Distant supervision: Performs distant supervision to combine learning and inference into a unified task.
  • MAP/marginal inference and weight learning: Support a set of tasks used in data analytics.