online_cvxNDL
This is the github repository of our Online Convex Network Dictionary Leanring algorithm (online cvxNDL).
Packages
The following packages are needed to run the code
python=3.8
pandas
matplotlib
ipdb
tqdm
scikit-learn
k-means-constrained
cvxpy
networkx
tensorly
psutil
To run the optimization algorithm
We provide synthetic data generated by the Stochastic Block Model (SBM) for quick verfication. For the full ChIA-Drop dataset, please refer to this link.
To run the example for the synthetic dataset
source run_example.sh
Note that we separate the MCMC sampling procedure from the iterative optimization procedure to improve the overall efficiency. The MCMC sampling procedure can be found in the folder sampling
.
To view the results in Jupyter notebook
Please go to the folder jupyter_notebook
and select the corresponding notebooks for detailed results.
Contact
Please contact Jianhao Peng (jianhao2@illinois.edu) or Chao Pan (chaopan2@illinois.edu) if you have any questions.