| The Adjacency Conservative PC Algorithm (ACPC) |
| Causal structure learning
in absence of faithfulness |
| This page will contain an applet with demos, the code
(Java, compatible with Tetrad) and references.
Unfortunately I had no time yet to finish it. Mail me if you are
interested. This will encourage me to finish it. |
Published in International Journal of Approximate Reasoning: Conservative Independence-Based Causal Structure Learning in Absence of Adjacency Faithfulness.
Based on: Jan Lemeire, Stijn Meganck, Francesco Cartella, Robust Independence-Based Causal Structure Learning in Absence of Adjacency Faithfulness, in Procs of European Workshop on Probabilistic Graphical Models (PGM), Helsinki, Finland, 2010.
This code is based on Tetrad 4.3.1, and an extension to the PC algorithm, which is fully explained in SGS '93 (Peter Spirtes, Clark Glymour and Richard Scheines. Causation, Prediction, and Search, Springer Verlag, 1993).
The applet is still under development...
If
necessary, install the latest
java-plugin (java 1.5 is required).
Alternatively, download
jar-file and run the module by double-clicking on it
or with command java
-jar causalLearningWithKde.jar (Under Windows:
Start => Run => execute cmd to open command shell, go to
correct directory with cd <directory>).
More information: Jan Lemeire (jan.lemeire@vub.ac.be)
See also Interactive
tutorial