Flexible Multilayer Sparse Approximations of Matrices and Applications
Authors:
- 17 March 2016
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Signal Processing
- Vol. 10 (4) , 688-700
Abstract
The computational cost of many signal processing and machine learning techniques is often dominated by the cost of applying certain linear operators to high-dimensional vectors. This paper introduces an algorithm aimed at reducing the complexity of applying linear operators in high dimension by approximately factorizing the corresponding matrix into few sparse factors. The approach relies on recent advances in nonconvex optimization. It is first explained and analyzed in details and then demonstrated experimentally on various problems including dictionary learning for image denoising and the approximation of large matrices arising in inverse problems.Keywords
All Related Versions
Funding Information
- European Research Council (ERC-StG- 2011-277906)
This publication has 36 references indexed in Scilit:
- Sample Complexity of Dictionary Learning and Other Matrix FactorizationsIEEE Transactions on Information Theory, 2015
- Chasing butterflies: In search of efficient dictionaries2015
- Learning Efficient Sparse and Low Rank ModelsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2015
- The Sparse Matrix Transform for Covariance Estimation and Analysis of High Dimensional SignalsIEEE Transactions on Image Processing, 2010
- Learning Deep Architectures for AIFoundations and Trends in Machine Learning, 2009
- Treelets—An adaptive multi-scale basis for sparse unordered dataThe Annals of Applied Statistics, 2008
- Reducing the Dimensionality of Data with Neural NetworksScience, 2006
- A Sparse Matrix Arithmetic Based on $\Cal H$ -Matrices. Part I: Introduction to ${\Cal H}$ -MatricesComputing, 1999
- Fast wavelet transforms and numerical algorithms ICommunications on Pure and Applied Mathematics, 1991
- Rapid solution of integral equations of classical potential theoryJournal of Computational Physics, 1985