Pseudo-likelihood (PL) bipartite clustering (+ spectral clustering).
Run test_real_data.m. The code implements the following:
biSpecClust3: Laplacian-type spectral clustering as discussed in the following paper: Matched Bipartite Block Model with Covariates, by Razaee, Amini and Li [1].biSpecClustDR: Ajdacency-based data-driven spectral clustering as discussed in the following paper: Analysis of spectral clustering algorithms for community detection: the general bipartite setting, by Zhou and Amini. This code is borrowed from the zhixin0825 /bipartite-spectral-clustering repo.PLEM: The PL-EM iterations for improving intial labels as discussed in the following paper: Optimal Bipartite Network Clustering, by Zhou and Amini.
The data is from [1]. The variational algorithm from [1] is implemented in aaamini/mbisbm repo., and it could improve the results if there are covariates on both sides and matched clustering is desired.