Joint Inference of Multiple Graphs with Hidden Variales
The code in this repository implements the algorithm introduced in the paper [1], submited to ICASSP 2022.
The organization of the repository is as follows:
- hidden_influence.m: includes the code of the test case 1 in the paper.
- samples_baselines.m: includes the code of the test case 2 in the paper.
- samples_real.m: includes the code of the test case 3 in the paper.
- opt: contains the optimization algorithms used in the previous simulations. Include the proposed algorithm and other alternatives used as baselines.
- utils: contains some utility functions such as generating similar ER graphs or scripts to perform a grid search for the weights of the regularizers.
- data: contains the dataset employed in the test case 3.
[1] S. Rey, A. Buciulea, M. Navarro, S. Segarra, A, Marques. "Joint inference of multiple graphs with hidden variables from stationary graph signals".