Sparse plus low rank network identification: a nonparametric approach
From MaRDI portal
Publication:503197
DOI10.1016/J.AUTOMATICA.2016.08.014zbMATH Open1352.93097arXiv1510.02961OpenAlexW2962875092MaRDI QIDQ503197FDOQ503197
Authors: Mattia Zorzi, Alessandro Chiuso 
Publication date: 11 January 2017
Published in: Automatica (Search for Journal in Brave)
Abstract: Modeling and identification of high-dimensional stochastic processes is ubiquitous in many fields. In particular, there is a growing interest in modeling stochastic processes with simple and interpretable structures. In many applications, such as econometrics and biomedical sciences, it seems natural to describe each component of that stochastic process in terms of few factor variables, which are not accessible for observation, and possibly of few other components of the stochastic process. These relations can be encoded in graphical way via a structured dynamic network, referred to as "sparse plus low-rank (S+L) network" hereafter. The problem of finding the S+L network as well as the dynamic model can be posed as a system identification problem. In this paper, we introduce two new nonparametric methods to identify dynamic models for stochastic processes described by a S+L network. These methods take inspiration from regularized estimators based on recently introduced kernels (e.g. "stable spline", "tuned-correlated" etc.). Numerical examples show the benefit to introduce the S+L structure in the identification procedure.
Full work available at URL: https://arxiv.org/abs/1510.02961
Recommendations
Gaussian processeskernel-based methodslinear system identificationsparsity and low-rank inducing priors
Stochastic network models in operations research (90B15) Identification in stochastic control theory (93E12)
Cites Work
- Estimating the dimension of a model
- Title not available (Why is that?)
- A new look at the statistical model identification
- Regularization algorithms for learning that are equivalent to multilayer networks
- Gaussian processes for machine learning.
- Theory of Reproducing Kernels
- Statistical analysis of network data. Methods and models
- Robust principal component analysis?
- 10.1162/15324430152748236
- Title not available (Why is that?)
- Rejoinder: Latent variable graphical model selection via convex optimization
- Investigating Causal Relations by Econometric Models and Cross-spectral Methods
- On the estimation of transfer functions, regularizations and Gaussian processes-revisited
- Maximum entropy properties of discrete-time first-order stable spline kernel
- Kernel methods in system identification, machine learning and function estimation: a survey
- Maximum Entropy Kernels for System Identification
- Prediction error identification of linear systems: a nonparametric Gaussian regression approach
- On covariance function tests used in system identification
- A new kernel-based approach for linear system identification
- Stochastic differential equations. An introduction with applications.
- A Bayesian approach to sparse dynamic network identification
- Convex vs non-convex estimators for regression and sparse estimation: the mean squared error properties of ARD and GLasso
- Axiomatic derivation of the principle of maximum entropy and the principle of minimum cross-entropy
- Iterated function systems, moments, and transformations of infinite matrices
- An Empirical Bayesian Strategy for Solving the Simultaneous Sparse Approximation Problem
- Granger causality: basic theory and application to neuroscience
- Quantifying the error in estimated transfer functions with application to model order selection
- AR Identification of Latent-Variable Graphical Models
- Multivariate Spectral Estimation Based on the Concept of Optimal Prediction
- A New Family of High-Resolution Multivariate Spectral Estimators
- On jointly stationary feedback-free stochastic processes
- System Identification by Dynamic Factor Models
- Gaussian Multiresolution Models: Exploiting Sparse Markov and Covariance Structure
- A scaled gradient projection method for Bayesian learning in dynamical systems
- A Maximum Entropy Enhancement for a Family of High-Resolution Spectral Estimators
- Modelling high-dimensional time series by generalized linear dynamic factor models: an introductory survey
Cited In (23)
- The harmonic analysis of kernel functions
- Sparse estimation in linear dynamic networks using the stable spline horseshoe prior
- High precision variational Bayesian inference of sparse linear networks
- The role of sparsity in inverse problems for networks with nonlinear dynamics
- Sparse network identifiability via compressed sensing
- Identification of low rank vector processes
- Maximum entropy vector kernels for MIMO system identification
- New reproducing kernel functions in the reproducing kernel Sobolev spaces
- Finite time identification in unstable linear systems
- Stationary-sparse causality network learning
- Topology identification of sparse network: a stochastic variational Bayesian approach
- A shift in paradigm for system identification
- Topology identification under spatially correlated noise
- A frequency domain approach for local module identification in dynamic networks
- A Bayesian approach to sparse dynamic network identification
- Generic identifiability of subnetworks in a linear dynamic network: the full measurement case
- Abstractions of linear dynamic networks for input selection in local module identification
- Sparse plus low-rank identification for dynamical latent-variable graphical AR models
- Identification of dynamical networks
- Support Recovery and Parameter Identification of Multivariate ARMA Systems with Exogenous Inputs
- Nonparametric identification of Kronecker networks
- An analysis of the SPARSEVA estimate for the finite sample data case
- Learning linear modules in a dynamic network using regularized kernel-based methods
This page was built for publication: Sparse plus low rank network identification: a nonparametric approach
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q503197)
