Year: 2007
Type: article
Abstract: latentnet is a package to fit and evaluate statistical latent position\n and cluster models for networks. Hoff, Raftery, and Handcock (2002)\n suggested an approach to modeling networks based on positing the existence\n of an latent space of characteristics of the actors. Relationships form as\n a function of distances between these characteristics as well as functions\n of observed dyadic level covariates. In latentnet social distances are\n represented in a Euclidean space. It also includes a variant of the\n extension of the latent position model to allow for clustering of the\n positions developed in Handcock, Raftery, and Tantrum (2007).\tThe package\n implements Bayesian inference for the models based on an Markov chain\n Monte Carlo algorithm. It can also compute maximum likelihood estimates\n for the latent position model and a two-stage maximum likelihood method\n for the latent position cluster model. For latent position cluster models,\n the package provides a Bayesian way of assessing how many groups there\n are, and thus whether or not there is any clustering (since if the\n preferred number of groups is 1, there is little evidence for clustering).\n It also estimates which cluster each actor belongs to. These estimates are\n probabilistic, and provide the probability of each actor belonging to each\n cluster. It computes four types of point estimates for the coefficients\n and positions: maximum likelihood estimate, posterior mean, posterior mode\n and the estimator which minimizes Kullback-Leibler divergence from the\n posterior. You can assess the goodness-of-fit of the model via posterior\n predictive checks. It has a function to simulate networks from a latent\n position or latent position cluster model.
Authors: Pavel N. Krivitsky, Mark S. Handcock
Language: English
Subfield: Statistical and Nonlinear Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
Funder: National Institute on Drug Abuse
OpenAlex ID: W4405774555