.png){kind=icon}
.png){kind=icon}
Return
Structure learning for weighted networks based on Bayesian nonparametric models
.png){kind=icon}
PRE
AI
DOI：10.1007/s13042-015-0439-1.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
With the increase of availability and scope of complex networks, structure learning for networks has received an enormous amount of interest in many fields, including physics, computer and information sciences, biology and the social sciences. To extract compact and flexible representations for weighted networks, we propose a new Bayesian nonparametric model to learn from both the existence and weight of interactions between nodes. Our model adopts Dirichlet process prior to automatically infer the partition over nodes in weighted networks without specifying the number of clusters. This is vital for structure discovery in complex networks, especially for novel domains where we have little prior knowledge. We develop a mean-field variational algorithm to efficiently approximate the model's posterior distribution over infinite latent clusters. Conducting extensive experiments on synthetic data set and four popular data sets, we demonstrate that our model can effectively capture the latent structure for complex weighted networks.
Keywords:
Structure learning
Clustering
Probabilistic graph models
Bayesian nonparametric models
Variational inference
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
2.7
Papers:
3.1K
Citations:
5.6K