Comparative Study for Inference of Hidden Classes in Stochastic Block Models
Machine Learning
2013-02-05 v2 Statistical Mechanics
Data Analysis, Statistics and Probability
Machine Learning
Abstract
Inference of hidden classes in stochastic block model is a classical problem with important applications. Most commonly used methods for this problem involve na\"{\i}ve mean field approaches or heuristic spectral methods. Recently, belief propagation was proposed for this problem. In this contribution we perform a comparative study between the three methods on synthetically created networks. We show that belief propagation shows much better performance when compared to na\"{\i}ve mean field and spectral approaches. This applies to accuracy, computational efficiency and the tendency to overfit the data.
Cite
@article{arxiv.1207.2328,
title = {Comparative Study for Inference of Hidden Classes in Stochastic Block Models},
author = {Pan Zhang and Florent Krzakala and Jörg Reichardt and Lenka Zdeborová},
journal= {arXiv preprint arXiv:1207.2328},
year = {2013}
}
Comments
8 pages, 5 figures AIGM12