Non-convex feature learning via l p, ∞ operator
Explore this paper's citation graph
Summary
A feature selection method is presented for solving sparse regularization problem, which has a composite regularization of lp norm and l∞ norm, and brings some insight for solving sparsity-favoring norm.
- Type
- article
- Published
- 2014-07-27
- Cited by
- 8
- References
- 31
- OpenAlex
- https://openalex.org/W1180959455
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:14320590
Keywords
Norm (philosophy), Proximal gradient methods for learning, Regularization (linguistics), Regular polygon, Feature selection
References
- Modeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope
- Selection of Relevant Features in Machine Learning
- Multi-Task Feature Learning Via Efficient l2, 1-Norm Minimization
- Asymptotic properties of bridge estimators in sparse high-dimensional regression models
- Robust nonnegative matrix factorization using L21-norm
- Wrappers for Feature Subset Selection
- Restricted isometry properties and nonconvex compressive sensing
- Efficient Methods for Overlapping Group Lasso
- Simultaneous Variable Selection
- The Elements of Statistical Learning
- Multi-label ReliefF and F-statistic feature selections for image annotation
- Algorithms for simultaneous sparse approximation. Part II: Convex relaxation
- Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties
- Efficient Algorithms for Selecting Features with Arbitrary Group Constraints via Group Lasso
- From Transformation-Based Dimensionality Reduction to Feature Selection
- Transfer learning for image classification with sparse prototype representations
- An Evaluation of Statistical Approaches to Text Categorization
- Structure learning in random fields for heart motion abnormality detection
- Introductory Lectures on Convex Optimization - A Basic Course
- Minimum redundancy feature selection from microarray gene expression data
Cited by
- Exclusive Feature Learning on Arbitrary Structures via ℓ_1, 2-norm
- Spectrally constrained L1-norm improves quantitative accuracy of diffuse optical tomography
- L1-norm based nonlinear reconstruction improves quantitative accuracy of spectral diffuse optical tomography
- Cost-Sensitive Feature Selection by Optimizing F-Measures
- Discriminative Feature Selection via Employing Smooth and Robust Hinge Loss
- Commercial Video Evaluation via Low-Level Feature Extraction and Selection
- University of Birmingham L1-norm Based Nonlinear Reconstruction Improves Quantitative Accuracy of Spectral Diffuse Optical Tomography
- University of Birmingham L1-norm Based Nonlinear Reconstruction Improves Quantitative Accuracy of Spectral Diffuse Optical Tomography
Related papers
- l_2,p Matrix Norm and Its Application in Feature Selection
- Primal-Dual Framework for Feature Selection using Least Squares Support Vector Machines
- Robust L1-norm multi-weight vector projection support vector machine with efficient algorithm
- Sparse subspace clustering via smoothed ℓp minimization
- Multi-Task Feature Learning Via Efficient l2,1-Norm Minimization
- Non-Convex Transfer Subspace Learning for Unsupervised Domain Adaptation
- Efficient inexact proximal gradient algorithms for structured sparsity-inducing norm
- Sparsity reconstruction using nonconvex TGpV-shearlet regularization and constrained projection
- Structured Sparsity via Alternating Direction Methods