/GLTM

Mind Foundry wrapper for the DataTypes packages

Primary LanguageC++

General Latent Type Model

(Automatic Discovery of the Statistical Types of Variables in a Dataset)

This code implements the Bayesian method and reproduces the experiment in

 I. Valera and Z. Ghahramani, 
 "Automatic Discovery of the Statistical Types of Variables in a Dataset", 
 34th International Conference on Machine Learning (ICML 2017). Sydney (Australia), 2017.

This version contains a private (MF owned) python wrapper that enables usage in python 3. This is a fairly extensive C++ re-write of the original code. To use the wrapper:

  1. Go to Ccode/wrapper_python
  2. Follow the instructions therein (includes installing various packages and whatnot)