r.blip: Bayesian Network Learning Improved Project

Allows the user to learn Bayesian networks from datasets containing thousands of variables. It focuses on score-based learning, mainly the 'BIC' and the 'BDeu' score functions. It provides state-of-the-art algorithms for the following tasks: (1) parent set identification - Mauro Scanagatta (2015) <http://papers.nips.cc/paper/5803-learning-bayesian-networks-with-thousands-of-variables>; (2) general structure optimization - Mauro Scanagatta (2018) <doi:10.1007/s10994-018-5701-9>, Mauro Scanagatta (2018) <http://proceedings.mlr.press/v73/scanagatta17a.html>; (3) bounded treewidth structure optimization - Mauro Scanagatta (2016) <http://papers.nips.cc/paper/6232-learning-treewidth-bounded-bayesian-networks-with-thousands-of-variables>; (4) structure learning on incomplete data sets - Mauro Scanagatta (2018) <doi:10.1016/j.ijar.2018.02.004>. Distributed under the LGPL-3 by IDSIA.

Getting started

Package details

AuthorMauro Scanagatta [aut, cre]
MaintainerMauro Scanagatta <mauro@idsia.ch>
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:

Try the r.blip package in your browser

Any scripts or data that you put into this service are public.

r.blip documentation built on May 2, 2019, 3:01 a.m.