Learning mixed kronecker product graph models with simulated method of moments

Sebastian Moreno, Jennifer Neville, Sergey Kirshner

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

13 Scopus citations

Abstract

There has recently been a great deal of work focused on de- veloping statistical models of graph structure|with the goal of modeling probability distributions over graphs from which new, similar graphs can be generated by sampling from the estimated distributions. Although current graph models can capture several important characteristics of social network graphs (e.g., degree, path lengths), many of them do not generate graphs with sufficient variation to reect the natural variability in real world graph domains. One exception is the mixed Kronecker Product Graph Model (mKPGM), a generalization of the Kronecker Product Graph Model, which uses parameter tying to capture variance in the underlying distribution [10]. The enhanced representation of mKPGMs enables them to match both the mean graph statistics and their spread as observed in real network populations, but un- fortunately to date, the only method to estimate mKPGMs involves an exhaustive search over the parameters. In this work, we present the first learning algorithm for mKPGMs. The O(jEj) algorithm searches over the contin- uous parameter space using constrained line search and is based on simulated method of moments, where the objective function minimizes the distance between the observed moments in the training graph and the empirically estimated moments of the model. We evaluate the mKPGM learning algorithm by comparing it to several different graph models, including KPGMs. We use multi-dimensional KS distance to compare the generated graphs to the observed graphs and the results show mKPGMs are able to produce a closer match to real-world graphs (10-90% reduction in KS distance), while still providing natural variation in the generated graphs.

Original languageEnglish
Title of host publicationKDD 2013 - 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
EditorsRajesh Parekh, Jingrui He, Dhillon S. Inderjit, Paul Bradley, Yehuda Koren, Rayid Ghani, Ted E. Senator, Robert L. Grossman, Ramasamy Uthurusamy
PublisherAssociation for Computing Machinery
Pages1052-1060
Number of pages9
ISBN (Electronic)9781450321747
DOIs
StatePublished - 11 Aug 2013
Externally publishedYes
Event19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2013 - Chicago, United States
Duration: 11 Aug 201314 Aug 2013

Publication series

NameProceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
VolumePart F128815

Conference

Conference19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD 2013
Country/TerritoryUnited States
CityChicago
Period11/08/1314/08/13

Keywords

  • Kronecker models
  • Link analysis
  • Method of moments estimation
  • Statistical graph models

Fingerprint

Dive into the research topics of 'Learning mixed kronecker product graph models with simulated method of moments'. Together they form a unique fingerprint.

Cite this