Two-Layer Feature Reduction for Sparse-Group Lasso via Decomposition of Convex Sets
作者： Jie Wang , Zhanqiu Zhang , Jieping Ye
DOI:
关键词: Dual (category theory) 、 Feature (computer vision) 、 Sparse approximation 、 Artificial intelligence 、 Regular polygon 、 Reduction (complexity) 、 Pattern recognition 、 Group (mathematics) 、 Lasso (statistics) 、 Feasible region 、 Mathematics
摘要: … dual feasible solutions of SGL via the framework of Fenchel’s duality (see Section 3). Based on the Fenchel’s dual problem of SGL, … , we analyze the dual feasible set of SGL in depth via …
nips.cc
arxiv.org
harvard.edu
elsevier.com
acm.org参考文章(28)
Shuiwang Ji, Jun Liu, Jieping Ye, SLEP: Sparse Learning with Efficient Projections ,(2011)
Jie Wang, Jun Liu, Jieping Ye, Efficient Mixed-Norm Regularization: Algorithms and Safe Screening Methods arXiv: Learning. ,(2013)
Kohei Ogawa, Yoshiki Suzuki, Ichiro Takeuchi, Safe Screening of Non-Support Vectors in Pathwise SVM Computation international conference on machine learning. pp. 1382- 1390 ,(2013)
Jonathan M. Borwein, Adrian S. Lewis, Convex analysis and nonlinear optimization ,(2006)
Jean-Baptiste Hiriart-Urruty, A Note on the Legendre-Fenchel Transform of Convex Composite Functions Springer, Boston, MA. pp. 35- 46 ,(2006) , 10.1007/0-387-29195-4_4
Patrick L. Combettes, Heinz H. Bauschke, Convex Analysis and Monotone Operator Theory in Hilbert Spaces ,(2011)
Tarek Rabbani, Laurent El Ghaoui, Vivian Viallon, Safe Feature Elimination in Sparse Supervised Learning arXiv: Learning. ,(2010)
Shinya Suzumura, Kohei Ogawa, Yoshiki Suzuki, Ichiro Takeuchi, Safe Sample Screening for Support Vector Machines arXiv: Machine Learning. ,(2014)
Jie Wang, Peter Wonka, Peter Wonka, Jieping Ye, Scaling SVM and Least Absolute Deviations via Exact Data Reduction international conference on machine learning. pp. 523- 531 ,(2014)
Noah Simon, Jerome Friedman, Trevor Hastie, Robert Tibshirani, A Sparse-Group Lasso Journal of Computational and Graphical Statistics. ,vol. 22, pp. 231- 245 ,(2013) , 10.1080/10618600.2012.681250