Adaptive First-Order Methods for General Sparse Inverse Covariance Selection
From MaRDI portal
Abstract: In this paper, we consider estimating sparse inverse covariance of a Gaussian graphical model whose conditional independence is assumed to be partially known. Similarly as in [5], we formulate it as an -norm penalized maximum likelihood estimation problem. Further, we propose an algorithm framework, and develop two first-order methods, that is, the adaptive spectral projected gradient (ASPG) method and the adaptive Nesterov's smooth (ANS) method, for solving this estimation problem. Finally, we compare the performance of these two methods on a set of randomly generated instances. Our computational results demonstrate that both methods are able to solve problems of size at least a thousand and number of constraints of nearly a half million within a reasonable amount of time, and the ASPG method generally outperforms the ANS method.
Recommendations
- First-Order Methods for Sparse Covariance Selection
- On ${l}_{q}$ Optimization and Sparse Inverse Covariance Selection
- Certifiably optimal sparse inverse covariance estimation
- Adaptive thresholding for sparse covariance matrix estimation
- Adaptive variable selection in nonparametric sparse regression
- <formula formulatype="inline"><tex Notation="TeX">$l_{0}$</tex></formula> Sparse Inverse Covariance Estimation
- Fast algorithms for sparse inverse covariance estimation
- Alternating direction method for covariance selection models
- Adaptive covariance estimation with model selection
- Smooth Optimization Approach for Sparse Covariance Selection
Cited in
(19)- Ridge estimation of inverse covariance matrices from high-dimensional data
- An efficient algorithm for sparse inverse covariance matrix estimation based on dual formulation
- A dual spectral projected gradient method for log-determinant semidefinite problems
- Sparse estimation of high-dimensional inverse covariance matrices with explicit eigenvalue constraints
- Alternating direction method for covariance selection models
- On how to solve large-scale log-determinant optimization problems
- QUIC: quadratic approximation for sparse inverse covariance estimation
- First-Order Methods for Sparse Covariance Selection
- Smooth Optimization Approach for Sparse Covariance Selection
- On the existence of the weighted bridge penalized Gaussian likelihood precision matrix estimator
- Positive-definite _1-penalized estimation of large covariance matrices
- scientific article; zbMATH DE number 7307482 (Why is no real title available?)
- Fused multiple graphical lasso
- Randomized block proximal damped Newton method for composite self-concordant minimization
- An inexact interior point method for \(L_{1}\)-regularized sparse covariance selection
- pISTA: preconditioned Iterative Soft Thresholding Algorithm for Graphical Lasso
- A block coordinate gradient descent method for regularized convex separable optimization and covariance selection
- A dual adaptive algorithm for matrix optimization with sparse group Lasso regularization
- A nonmonotone active-set semismooth Newton method for matrix approximation with group regularization
This page was built for publication: Adaptive First-Order Methods for General Sparse Inverse Covariance Selection
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3053132)
