https://github.com/ax-va/project-bayesianpy

This is a collection of algorithms and models written in Python for probabilistic programming. The main focus of the package is on Bayesian reasoning by using Bayesian networks, Markov networks, and their mixing.
https://github.com/ax-va/project-bayesianpy

bayesian-inference bayesian-machine-learning bayesian-network bayesian-networks bayesian-reasoning inference machine-learning machine-learning-algorithms markov-network markov-networks pgm probabilistic-graphical-models probabilistic-programming probabilistic-reasoning

Last synced: about 18 hours ago
JSON representation

This is a collection of algorithms and models written in Python for probabilistic programming. The main focus of the package is on Bayesian reasoning by using Bayesian networks, Markov networks, and their mixing.

• Host: GitHub
• URL: https://github.com/ax-va/project-bayesianpy
• Owner: ax-va
• License: bsd-3-clause
• Created: 2021-09-09T06:49:15.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2023-04-29T21:18:59.000Z (over 1 year ago)
• Last Synced: 2023-04-29T22:58:58.916Z (over 1 year ago)
• Topics: bayesian-inference, bayesian-machine-learning, bayesian-network, bayesian-networks, bayesian-reasoning, inference, machine-learning, machine-learning-algorithms, markov-network, markov-networks, pgm, probabilistic-graphical-models, probabilistic-programming, probabilistic-reasoning
• Language: Python
• Homepage:
• Size: 1.54 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 6
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# BayesianPy

This is a collection of algorithms and models written in Python 3.8 for probabilistic programming. The main focus of the package is on Bayesian reasoning by using Bayesian networks, Markov networks, and their mixing.

The collection contains the following algorithms and models.

Factored-inference-related algorithms for probabilistic graphical models with categorical distributions:

- Belief Propagation (BP) [B12] for efficient inference in trees (pb4ml/inference/factored/belief_propagation.py);

- Bucket Elimination (BE) [B12] for inference in loopy graphs or computing the joint probability distribution of several query variables (pb4ml/inference/factored/bucket_elimination.py);

- Greedy Bucket Elimination (GBE) combining Bucket Elimination with an elimination ordering pre-calculated by Greedy Ordering (pb4ml/inference/factored/greedy_elimination.py);

- Greedy Ordering (GO) [KF09] for greedy search for a near-optimal variable elimination ordering (pb4ml/inference/factored/greedy_ordering.py).

Academic probabilistic models in the factor graph representation:

- Bayesian network "Extended Student" [KF09] (pb4ml/models/academic/extended_student.py);

- Bayesian network "Student" [KF09] (pb4ml/models/academic/student.py);

- Markov network "Misconception" [KF09] (pb4ml/models/academic/misconception.py).

See in the tests folder how to use the algorithms. In the models folder, you can see how to create factor graph models.

© 2021-2023 Alexander Vasiliev

References:

[B12] David Barber, "Bayesian Reasoning and Machine Learning", Cambridge University Press, 2012;

[KF09] Daphne Koller and Nir Friedman, "Probabilistic Graphical Models: Principles and Techniques", The MIT Press, 2009