Info-Clustering
Clustering random variables by Multivariate Mutual Information (MMI)
|
The above binder link launches a notebook that demonstrates the agglomerative info-clustering algorithm (AIC) implemented in C++. It is run using the xeus-cling C++11 jupyter kernel.
Simply download the include
folder and add it to the include path.
include
folder.libpari-dev
. (See the next section.)Some examples from the jupyter notebook are also included as .cpp
files as follows. See the jupyter notebook for more detailed explanations.
hypergraph_demo.cpp performs the exact and approximate clustering algorithm for hypergraphical source model. Run using
gaussian_demo.cpp gives an example of a jointly gaussian source model. Run using
fls_demo.cpp is an example for clustering a finite linear source model. It requires the C library pari, which can be installed on ubuntu using
Run the using