Alternating proximal gradient method for sparse nonnegative Tucker decomposition
From MaRDI portal
Publication:2356334
alternating proximal gradient methodnon-convex optimizationsparse nonnegative Tucker decompositionsparse optimization
Numerical methods of relaxation type (49M20) Decomposition methods (49M27) Numerical methods based on nonlinear programming (49M37) Extrapolation to the limit, deferred corrections (65B05) Nonconvex programming, global optimization (90C26) Nonlinear programming (90C30) Methods of reduced gradient type (90C52)
Abstract: Multi-way data arises in many applications such as electroencephalography (EEG) classification, face recognition, text mining and hyperspectral data analysis. Tensor decomposition has been commonly used to find the hidden factors and elicit the intrinsic structures of the multi-way data. This paper considers sparse nonnegative Tucker decomposition (NTD), which is to decompose a given tensor into the product of a core tensor and several factor matrices with sparsity and nonnegativity constraints. An alternating proximal gradient method (APG) is applied to solve the problem. The algorithm is then modified to sparse NTD with missing values. Per-iteration cost of the algorithm is estimated scalable about the data size, and global convergence is established under fairly loose conditions. Numerical experiments on both synthetic and real world data demonstrate its superiority over a few state-of-the-art methods for (sparse) NTD from partial and/or full observations. The MATLAB code along with demos are accessible from the author's homepage.
Recommendations
Cites work
- A block coordinate descent method for regularized multiconvex optimization with applications to nonnegative tensor factorization and completion
- A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
- Algorithms for Sparse Nonnegative Tucker Decompositions
- An alternating direction algorithm for matrix completion with nonnegative factors
- Analysis of individual differences in multidimensional scaling via an \(n\)-way generalization of ``Eckart-Young decomposition
- Computing non-negative tensor factorizations
- Exploiting symmetry in tensors for high performance: multiplication with symmetric tensors
- scientific article; zbMATH DE number 194139 (Why is no real title available?)
- Joint orthomax rotation of the core and component matrices resulting from three-mode principal components analysis
- Learning the parts of objects by non-negative matrix factorization
- Nonnegative Matrix Factorization Based on Alternating Nonnegativity Constrained Least Squares and Active Set Method
- On semi- and subanalytic geometry
- On the Best Rank-1 and Rank-(R1 ,R2 ,. . .,RN) Approximation of Higher-Order Tensors
- Solving a low-rank factorization model for matrix completion by a nonlinear successive over-relaxation algorithm
- Sparse non-negative tensor factorization using columnwise coordinate descent
- Tensor Decompositions and Applications
- The Łojasiewicz Inequality for Nonsmooth Subanalytic Functions with Applications to Subgradient Dynamical Systems
Cited in
(29)- A globally convergent algorithm for nonconvex optimization based on block coordinate update
- Structured nonconvex and nonsmooth optimization: algorithms and iteration complexity analysis
- Parameter selection for nonnegative $l_1$ matrix/tensor sparse decomposition
- Nonnegative tensor factorizations using an alternating direction method
- A primal-dual algorithm for nonnegative N-th order CP tensor decomposition: application to fluorescence spectroscopy data analysis
- Primal-dual optimization algorithms over Riemannian manifolds: an iteration complexity analysis
- Unsupervised machine learning based on non-negative tensor factorization for analyzing reactive-mixing
- Global and local structure preserving sparse subspace learning: an iterative approach to unsupervised feature selection
- Fast nonnegative tensor factorizations with tensor train model
- T-product factorization based method for matrix and tensor completion problems
- Algorithms for Sparse Nonnegative Tucker Decompositions
- Bregman iteration algorithm for sparse nonnegative matrix factorizations via alternating l₁-norm minimization
- Nonnegative tensor decomposition with custom clustering for microphase separation of block copolymers
- Orthogonal nonnegative Tucker decomposition
- Low rank non-negative triple decomposition and non-negative tensor completion
- Alternating structure-adapted proximal gradient descent for nonconvex nonsmooth block-regularized problems
- Nonnegative tensor decomposition
- Triple decomposition and tensor recovery of third order tensors
- Correntropy based model predictive controller with multi-constraints for robust path trajectory tracking of self-driving vehicle
- Study of performance of low-rank nonnegative tensor factorization methods
- Convergence of Bregman Peaceman-Rachford splitting method for nonconvex nonseparable optimization
- Sparse non-negative tensor factorization using columnwise coordinate descent
- Generalized nonconvex hyperspectral anomaly detection via background representation learning with dictionary constraint
- A partial Bregman ADMM with a general relaxation factor for structured nonconvex and nonsmooth optimization
- Alternating linearized proximal algorithm for the low Tucker rank tensor completion problem
- Tensor factorization-based method for tensor completion with spatio-temporal characterization
- Coordinate-update algorithms can efficiently detect infeasible optimization problems
- Blind hyperspectral and multispectral images fusion: a unified tensor fusion framework from coupled inverse problem perspective
- Hypergraph regularized and low-rank enhanced Tucker decomposition for image restoration
This page was built for publication: Alternating proximal gradient method for sparse nonnegative Tucker decomposition
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2356334)
