ahmadfaizalbh / pypnl Goto Github PK

PyPNL is a probabilistic network library in Python. The library consists of a Python API which interfaces to state of the art C/C++ inference and learning engines. The API design is modular and allows for plug-ins which augment functionality. PyPNL comes with documentation, examples, and a test suite.

The PyPNL library has a rich set of features for representation, inference, learning and actions using probabilistic networks.

• Discrete nodes
• Continuous nodes
• Utility nodes
• Decision nodes
• Static Bayesian networks
• Dynamic Bayesian networks
• Markov random fields
• Limited memory influence diagrams

• Exact enumeration
• Junction Tree
• Pearl
• Gibbs sampling

• K2
• Chow-Liu

• Decision nodes

Stay tuned!

Stay tuned!

Stay tuned!