Author: Kanvaly Fadiga
This repository implements our causal relation learning algorithm and also allows us to explain events. To facilitate its use we have used streamlit which allows us to create a web interface with which we can interact.
my python version: Python 3.8.5
after activating your python environment, install required package using this command:
pip install -r requirements.txt-
Web app: To run the web app run
python main.py -
Simple example: we also provide a simple python code
example.pythat learn a simple network. just runpython example.pyand it learn the graph and plot the difference with the groundthruth.
- CausalLeaner in
causal_learner.py:
contains the algorithm that learn causal relationship using intervention and interaction.
- BayesianNetwork in
network.py:
Our implementation og bayesian network. it contain function to perform do operation and data generation
- most_probable_explanation and belief_propagation in
explanation.py:
perform the two methods to do abduction in a bayesian network