Download this Paper Open PDF in Browser

Inference on the Effects of Observed Features in Latent Space Models for Networks

18 Pages Posted: 21 Jul 2016  

Zachary M. Jones

Pennsylvania State University

Matthew James Denny

Pennsylvania State University

Bruce A. Desmarais

Pennsylvania State University

Hanna Wallach

Microsoft Research New York City

Date Written: July 20, 2016

Abstract

The latent space model (LSM) for network data is a generative probabilistic model that combines a generalized linear model with a latent spatial embedding of the network. It has been used to decrease error in the estimation of and inference regarding the effects of observed covariates. In applications of the LSM, it is assumed that the latent spatial embedding can control for unmeasured confounding structure that is related to the values of edges in the network. As far as we know, there has been no research that considers the LSM’s performance in adjusting for unmeasured structure to reduce estimation and inferential errors. We investigate the LSM’s performance via a Monte Carlo study. In the presence of an unmeasured covariate that can be appropriately modeled using a latent space, estimation and inferential error remain high under even moderate confounding. However, the prediction error of the LSM when unmeasured network structure is present is substantially lower in most cases. We conclude that the LSM is most appropriately used for exploratory or predictive tasks.

Suggested Citation

Jones, Zachary M. and Denny, Matthew James and Desmarais, Bruce A. and Wallach, Hanna, Inference on the Effects of Observed Features in Latent Space Models for Networks (July 20, 2016). Available at SSRN: https://ssrn.com/abstract=2812240 or http://dx.doi.org/10.2139/ssrn.2812240

Zachary M. Jones (Contact Author)

Pennsylvania State University ( email )

University Park
State College, PA 16802
United States

Matthew James Denny

Pennsylvania State University ( email )

Bruce A. Desmarais

Pennsylvania State University ( email )

University Park, State College, PA 16801
United States

HOME PAGE: http://sites.psu.edu/desmaraisgroup

Hanna Wallach

Microsoft Research New York City ( email )

641 Avenue of Americas
New York, NY 10011
United States

Paper statistics

Downloads
40
Abstract Views
269