Title :
Fitting Network Data Based on Latent Cluster Model
Author :
Guo, Ying ; Wang, Xuefeng ; Zhu, Donghua ; Zhou, Xiao
Author_Institution :
Sch. of Manage. & Econ., Beijing Inst. of Technol., Beijing, China
Abstract :
In the last ten years, social network analysis became a very popular topic in many different scientific fields, network models are also widely popular for representing the relationship of the network data. Network data exhibits transitivity and homophily of the actors. There exist many distance computation methods for the actors space distance, and two of them are the most famous for the latent position cluster model, here we used the latent cluster model which focus on clusters of actors or ties. In this paper, we compared two distance definition methods with different latent position cluster method, the two-stage method with Euclidean distance(TMED) model and the bayesian estimation method with Bilinear latent(BEBL) model. The model make simulate the network dataset easy, and compared the mcmc diagnostics.
Keywords :
Bayes methods; pattern clustering; social networking (online); social sciences computing; Bayesian estimation method; Euclidean distance; actor homophily; actor space distance; actor transitivity; bilinear latent model; distance computation method; distance definition method; latent position cluster model; network data fitting; network data relationship; network model; social network analysis; Bayesian methods; Computational modeling; Data models; Educational institutions; Maximum likelihood estimation; Predictive models; Social network services;
Conference_Titel :
Management and Service Science (MASS), 2011 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-6579-8
DOI :
10.1109/ICMSS.2011.5999182