Selection of sparse vine copulas in high dimensions with the Lasso
From MaRDI portal
Publication:2329765
DOI10.1007/S11222-018-9807-5zbMATH Open1430.62102arXiv1705.05877OpenAlexW2616758160WikidataQ130069089 ScholiaQ130069089MaRDI QIDQ2329765FDOQ2329765
Authors: Dominik Müller, Claudia Czado 
Publication date: 18 October 2019
Published in: Statistics and Computing (Search for Journal in Brave)
Abstract: We propose a novel structure selection method for high dimensional (d > 100) sparse vine copulas. Current sequential greedy approaches for structure selection require calculating spanning trees in hundreds of dimensions and fitting the pair copulas and their parameters iteratively throughout the structure selection process. Our method uses a connection between the vine and structural equation models (SEMs). The later can be estimated very fast using the Lasso, also in very high dimensions, to obtain sparse models. Thus, we obtain a structure estimate independently of the chosen pair copulas and parameters. Additionally, we define the novel concept of regularization paths for R-vine matrices. It relates sparsity of the vine copula model in terms of independence copulas to a penalization coefficient in the structural equation models. We illustrate our approach and provide many numerical examples. These include simulations and data applications in high dimensions, showing the superiority of our approach to other existing methods.
Full work available at URL: https://arxiv.org/abs/1705.05877
Recommendations
- Model selection in sparse high-dimensional vine copula models with an application to portfolio risk
- Dependence modelling in ultra high dimensions with vine copulas and the graphical Lasso
- Vine copula models with GLM and sparsity
- The sparsity and bias of the LASSO selection in high-dimensional linear regression
- Necessary and sufficient conditions for variable selection consistency of the Lasso in high dimensions
- Sparse covariance thresholding for high-dimensional variable selection
- High-dimensional graphs and variable selection with the Lasso
- High-dimensional variable selection with sparse random projections: measurement sparsity and statistical efficiency
- Variable selection in high-dimensional sparse multiresponse linear regression models
- Simultaneous Variable and Covariance Selection With the Multivariate Spike-and-Slab LASSO
Characterization and structure theory for multivariate probability distributions; copulas (62H05) Ridge regression; shrinkage estimators (Lasso) (62J07)
Cites Work
- Pair-copula constructions of multiple dependence
- Selecting and estimating regular vine copulae and application to financial returns
- Estimating the dimension of a model
- Dependence modeling with copulas
- Title not available (Why is that?)
- Title not available (Why is that?)
- Parsimonious parameterization of correlation matrices using truncated vines and factor analysis
- Probability density decomposition for conditionally dependent random variables modeled by vines
- High-dimensional graphs and variable selection with the Lasso
- Truncated regular vines in high dimensions with application to financial data
- Title not available (Why is that?)
- Uncertainty Analysis with High Dimensional Dependence Modelling
- Title not available (Why is that?)
- Regularization and Variable Selection Via the Elastic Net
- Sparse inverse covariance estimation with the graphical lasso
- Vines -- a new graphical model for dependent random variables.
- Simplified pair copula constructions -- limitations and extensions
- Title not available (Why is that?)
- Probabilistic graphical models.
- Sequential Bayesian model selection of regular vine copulas
- Normal linear regression models with recursive graphical Markov structure
- Title not available (Why is that?)
- Representing Sparse Gaussian DAGs as Sparse R-Vines Allowing for Non-Gaussian Dependence
Cited In (11)
- Copula shrinkage and portfolio allocation in ultra-high dimensions
- Model selection for discrete regular vine copulas
- Model selection in sparse high-dimensional vine copula models with an application to portfolio risk
- Dependence modelling in ultra high dimensions with vine copulas and the graphical Lasso
- Copulae: an overview and recent developments
- Regular vines with strongly chordal pattern of (conditional) independence
- Pair-copula models for analyzing family data
- Dependence structure estimation using copula recursive trees
- Truncation of vine copulas using fit indices
- Vine copula models with GLM and sparsity
- Sequential truncation of \(R\)-vine copula mixture model for high-dimensional datasets
Uses Software
This page was built for publication: Selection of sparse vine copulas in high dimensions with the Lasso
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2329765)
