Projection algorithms for nonconvex minimization with application to sparse principal component analysis
From MaRDI portal
Publication:312468
DOI10.1007/S10898-016-0402-ZzbMATH Open1353.90118OpenAlexW2299651280MaRDI QIDQ312468FDOQ312468
Authors: William Hager, Dzung T. Phan, Jiajie Zhu 
Publication date: 15 September 2016
Published in: Journal of Global Optimization (Search for Journal in Brave)
Abstract: We consider concave minimization problems over non-convex sets.Optimization problems with this structure arise in sparse principal component analysis. We analyze both a gradient projection algorithm and an approximate Newton algorithm where the Hessian approximation is a multiple of the identity. Convergence results are established. In numerical experiments arising in sparse principal component analysis, it is seen that the performance of the gradient projection algorithm is very similar to that of the truncated power method and the generalized power method. In some cases, the approximate Newton algorithm with a Barzilai-Borwein (BB) Hessian approximation can be substantially faster than the other algorithms, and can converge to a better solution.
Full work available at URL: https://arxiv.org/abs/1404.4132
Recommendations
- Generalized power method for sparse principal component analysis
- Projection sparse principal component analysis: an efficient least squares method
- An exact approach to sparse principal component analysis
- An augmented Lagrangian approach for sparse principal component analysis
- Sparse PCA: convex relaxations, algorithms and applications
nonconvex minimizationsparse principal component analysisBarzilai-Borwein methodgradient projectionapproximate Newton
Cites Work
- Probing the Pareto frontier for basis pursuit solutions
- Least angle regression. (With discussion)
- Atomic Decomposition by Basis Pursuit
- Generalized power method for sparse principal component analysis
- Optimal solutions for sparse principal component analysis
- Truncated power method for sparse eigenvalue problems
- Convex approximations to sparse PCA via Lagrangian duality
- Two-Point Step Size Gradient Methods
- Mirror descent and nonlinear projected subgradient methods for convex optimization.
- Convex Analysis
- A majorization-minimization approach to the sparse generalized eigenvalue problem
- Compressed sensing
- Sparse Reconstruction by Separable Approximation
- A Nonmonotone Line Search Technique for Newton’s Method
- Detecting high log-densities, an \(O(n^{1/4})\) approximation for densest \(k\)-subgraph
- Coresets, sparse greedy approximation, and the Frank-Wolfe algorithm
- On Finding Dense Subgraphs
- Simultaneous pursuit of out-of-sample performance and sparsity in index tracking portfolios
- A New Active Set Algorithm for Box Constrained Optimization
- Conditional gradient algorithms for rank-one matrix approximations with a sparsity constraint
- Approximation of dense-\(n/2\)-subgraph and the complement of min-bisection
- Projected Newton Methods for Optimization Problems with Simple Constraints
Cited In (17)
- Projected gradient approach to the numerical solution of the SCoTLASS
- On the robust PCA and Weiszfeld's algorithm
- Alternating Projections and Douglas-Rachford for Sparse Affine Feasibility
- An efficient algorithm for Fantope-constrained sparse principal subspace estimation problem
- A generalized inertial proximal alternating linearized minimization method for nonconvex nonsmooth problems
- On the Solution of ℓ0-Constrained Sparse Inverse Covariance Estimation Problems
- An active-set proximal quasi-Newton algorithm for ℓ1-regularized minimization over a sphere constraint
- A customized proximal point algorithm for stable principal component pursuit with nonnegative constraint
- Non-Negative Principal Component Analysis: Message Passing Algorithms and Sharp Asymptotics
- A Lagrange–Newton algorithm for tensor sparse principal component analysis
- Proximal Methods for Sparse Optimal Scoring and Discriminant Analysis
- Derivatives of probability functions: unions of polyhedra and elliptical distributions
- Low-complexity \(l_0\)-norm penalized shrinkage linear and widely linear affine projection algorithms
- Sparse Principal Component Analysis via Axis-Aligned Random Projections
- Linear Convergence of a Proximal Alternating Minimization Method with Extrapolation for \(\boldsymbol{\ell_1}\) -Norm Principal Component Analysis
- Stochastic proximal gradient method FOR \(\ell_1\) regularized optimization over a sphere
- Random projections for the nonnegative least-squares problem
Uses Software
This page was built for publication: Projection algorithms for nonconvex minimization with application to sparse principal component analysis
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q312468)
