Exploration of Large Networks with Covariates via Fast and Universal Latent Space Model Fitting
arXiv:1705.02372
Abstract
Latent space models are effective tools for statistical modeling and exploration of network data. These models can effectively model real world network characteristics such as degree heterogeneity, transitivity, homophily, etc. Due to their close connection to generalized linear models, it is also natural to incorporate covariate information in them. The current paper presents two universal fitting algorithms for networks with edge covariates: one based on nuclear norm penalization and the other based on projected gradient descent. Both algorithms are motivated by maximizing likelihood for a special class of inner-product models while working simultaneously for a wide range of different latent space models, such as distance models, which allow latent vectors to affect edge formation in flexible ways. These fitting methods, especially the one based on projected gradient descent, are fast and scalable to large networks. We obtain their rates of convergence for both inner-product models and beyond. The effectiveness of the modeling approach and fitting algorithms is demonstrated on five real world network datasets for different statistical tasks, including community detection with and without edge covariates, and network assisted learning.
References in corpus (21)
- Hyperbolic Geometry of Complex Networks
- Social Structure of Facebook Networks
- Matrix estimation by Universal Singular Value Thresholding
- Guaranteed Matrix Completion via Non-convex Factorization
- Sparse principal component analysis and iterative thresholding
- Fast community detection by SCORE
- Modeling homophily and stochastic equivalence in symmetric relational data
- Fast low-rank estimation by projected gradient descent: General statistical and algorithmic guarantees
- Rate-optimal graphon estimation
- Matrix Completion has No Spurious Local Minimum
- Low-rank Solutions of Linear Matrix Equations via Procrustes Flow
- Nonparametric graphon estimation
- Stochastic blockmodel approximation of a graphon: Theory and consistent estimation
- Convergence Analysis for Rectangular Matrix Completion Using Burer-Monteiro Factorization and Gradient Descent
- Spectral properties of distance matrices
- Optimal Estimation and Completion of Matrices with Biclustering Structures
- Achieving Optimal Misclassification Proportion in Stochastic Block Model
- Detecting Overlapping Communities in Networks Using Spectral Methods
- Geometric Network Comparison
- Community Detection in Networks with Node Features
- Generalized linear models with low rank effects for network data
Cited by in corpus (7)
- Minimax Rates in Network Analysis: Graphon Estimation, Community Detection and Hypothesis Testing
- Dual regularized Laplacian spectral clustering methods on community detection
- An theory of PCA and spectral clustering
- A Time-Varying Network for Cryptocurrencies
- A Latent Space Model for Multilayer Network Data
- An improved spectral clustering method for community detection under the degree-corrected stochastic blockmodel
- Factor Analysis on Citation, Using a Combined Latent and Logistic Regression Model