Colored Maximum Variance Unfolding

Summary

It is shown that MVU also optimizes a statistical dependence measure which aims to retain the identity of individual observations under the distance-preserving constraints, which allows for "colored" variants of MVU, which produce low-dimensional representations for a given task.

Keywords

Variance (accounting), Computer science, Heuristic, Representation (politics), Dimensionality reduction

References

Learning a kernel matrix for nonlinear dimensionality reductionKilian Q. Weinberger, Fei Sha, Lawrence K. Saul — 2004
The Fastest Mixing Markov Process on a Graph and a Connection to a Maximum Variance Unfolding ProblemJun Sun, Stephen Boyd, Lin Xiao — 2006
Supervised feature selection via dependence estimationLe Song, Alex Smola, Arthur Gretton — 2007
A dependence maximization view of clusteringLe Song, Alex Smola, Arthur Gretton — 2007
Neighbourhood Components AnalysisJacob Goldberger, Geoffrey E. Hinton, Sam T. Roweis — 2004
Dimensionality Reduction for Supervised Learning with Reproducing Kernel Hilbert SpacesKenji Fukumizu, Francis R. Bach, Michael I. Jordan — 2004
Graph Laplacian Regularization for Large-Scale Semidefinite ProgrammingB. Schölkopf, J. Platt, T. Hofmann — 2007
Measuring Statistical Dependence with Hilbert-Schmidt NormsA. Gretton, O. Bousquet, Alex Smola — 2005
Graph Laplacian Regularization for Large-Scale Semidefinite ProgrammingKilian Q. Weinberger, Fei Sha, Qihui Zhu — 2006

Cited by

Advances in dissimilarity-based data visualisationAndrej Gisbrecht — 2015
Feature-based transfer learning with real-world applicationsJialin Pan — 2010
Canonical locality preserving Latent Variable Model for discriminative pose inferenceYan Tian, Leonid Sigal, Fernando De la Torre — 2013
Modeling of mutual dependenciesArto Klami — 2008
Domain Adaptation via Transfer Component AnalysisSinno Jialin Pan, Ivor W. Tsang, James T. Kwok — 2009
Nonlinear Dynamic Field Embedding: On Hyperspectral Scene VisualizationDalton Lunga, Okan Erosy — 2012
Gradient-based kernel dimension reduction for supervised learningKenji Fukumizu, Chenlei Leng — 2011
Multilabel dimensionality reduction via dependence maximizationYin Zhang, Zhi‐Hua Zhou — 2008
Generalized sparse metric learning with relative comparisonsKaizhu Huang, Yiming Ying, Colin Campbell — 2011
Adaptive Dissimilarity Measures, Dimension Reduction and Visualization (University of Groningen)Kerstin Bunte — 2012
Ideal regularization for learning kernels from labelsBinbin Pan, Jianhuang Lai, Lixin Shen — 2014
Non-parametric kernel ranking approach for social image retrievalJinfeng Zhuang, Steven C. H. Hoi — 2010
Supervised principal component analysis: Visualization, classification and regression on subspaces and submanifoldsElnaz Barshan, Ali Ghodsi, Zohreh Azimifar — 2011
Adaptive weighted learning for linear regression problems via Kullback-Leibler divergenceZhizheng Liang, Youfu Li, Shixiong Xia — 2013
Gradient-Based Kernel Dimension Reduction for RegressionKenji Fukumizu, Chenlei Leng — 2014
Dimensionality reduction by Mixed Kernel Canonical Correlation AnalysisXiaofeng Zhu, Zi Huang, Heng Tao Shen — 2012
Scalable Nonparametric Low-Rank Kernel Learning Using Block Coordinate DescentEnliang Hu, James T. Kwok — 2015
Hyperparameter Selection in Kernel Principal Component AnalysisMd Ashad Alam — 2014
Data visualization by nonlinear dimensionality reductionAndrej Gisbrecht, Barbara Hammer — 2015
Geometry-aware metric learningZhengdong Lu, Prateek Jain, Inderjit S. Dhillon — 2009

Related papers

Learning a kernel matrix for nonlinear dimensionality reductionKilian Q. Weinberger, Fei Sha, Lawrence K. Saul — 2004
A global geometric framework for nonlinear dimensionality reduction.Joshua B. Tenenbaum, Vin de Silva, John Langford — 2000
Nonlinear dimensionality reduction by locally linear embedding.Sam T. Roweis, Lawrence K. Saul — 2000
Nonlinear Component Analysis as a Kernel Eigenvalue ProblemBernhard Schölkopf, Alexander J. Smola, Klaus‐Robert Müller — 1998
Stochastic Neighbor EmbeddingGeoffrey E. Hinton, Sam T. Roweis — 2002
Distance Metric Learning with Application to Clustering with Side-InformationEric P. Xing, Michael I. Jordan, Stuart Russell — 2002
Neighbourhood Components AnalysisJacob Goldberger, Geoffrey E. Hinton, Sam T. Roweis — 2004
THE USE OF MULTIPLE MEASUREMENTS IN TAXONOMIC PROBLEMSRonald Aylmer Fisher — 1936