BayesianNetwork: Interactive Bayesian Network Modeling and Analysis
Paul Govan · The Journal of Open Source Software · 2018
BayesianNetwork (Govan 2017) is a shiny (Chang et al. 2017) web application for Bayesian Network modeling and analysis, providing a front-end to the bnlearn (Scutari 2009) package for Bayesian Network learning.The application includes structural learning algorithms for learning the structure of the network with support for both discrete and continuous variables, parameter learning methods for estimating the network parameters, procedures for adding evidence to the network and performing Bayesian inference, and node and network utilities for measuring the importance of connections in the network.BayesianNetwork originated as a research project for risk analysis applications, where the primary motivation was to develop an app for modeling and analyzing Bayesian Networks in an interactive environment (Govan and Damnjanovic 2016).The goal is that the package be both powerful and intuitive, serving as a tool for researchers, educators and students alike.The package documentation includes examples and additional resources for running the software.