This repository contains BANSHEE - A MATLAB Toolbox for Non-Parametric Bayesian Networks.
The code is an update of the original version supporting SoftwareX paper: https://doi.org/10.1016/j.softx.2020.100588
Bayesian Networks (BNs) are probabilistic, graphical models for representing complex dependency structures. They have many applications in science and engineering. This toolbox implements a particularly powerful variant Non-Parametric BNs. The software allows for quantifying the BN, validating the underlying assumptions of the model, visualizing the network and its corresponding rank correlation matrix, and finally making inference with a BN based on existing or new evidence. The toolbox also includes some applied BN models published in recent scientific literature.
The src/ directory contains BANSHEE.mltbx file containing installer of the MATLAB toolbox.
The docs/ directory contains a quick start guide to the toolbox. Please consult the guide before using the toolbox.
This version (1.3) compared with published version supporting the SoftwareX publication (v1.1 and v1.2) has the following changes:
- Adds two new real-life models for predicting flood losses for the residential and commercial sectors;
- Updates the quick start guide;
- Corrects the description in "predict_coastal_erosion.m";
- Removes three .mat files which contained data alreary present in other .mat files;
- Updates references to the SoftwareX paper and other publications.
- The user can work with nodes as parametric distribution
- The diferent sample samples sizes to compute the d-calibration score
- Sample base conditioning
- Conditional and un-conditional histograms