论文信息 - Global Versus Local Methods in Nonlinear Dimensionality Reduction

Global Versus Local Methods in Nonlinear Dimensionality Reduction

Recently proposed algorithms for nonlinear dimensionality reduction fall broadly into two categories which have different advantages and disadvantages: global (Isomap [1]), and local (Locally Linear Embedding [2], Laplacian Eigenmaps [3]). We present two variants of Isomap which combine the advantages of the global approach with what have previously been exclusive advantages of local methods: computational sparsity and the ability to invert conformal maps.

Joshua B. Tenenbaum | Vin de Silva | J. Tenenbaum | V. Silva

[1] Joseph L. Zinnes,et al. Theory and Methods of Scaling. , 1958 .

[2] Teuvo Kohonen,et al. Self-Organization and Associative Memory , 1988 .

[3] Michael I. Jordan,et al. Advances in Neural Information Processing Systems 30 , 1995 .

[4] Garrison W. Cottrell,et al. Non-Linear Dimensionality Reduction , 1992, NIPS.

[5] Stephen M. Omohundro,et al. Nonlinear Image Interpolation using Manifold Learning , 1994, NIPS.

[6] Joshua B. Tenenbaum,et al. Mapping a Manifold of Perceptual Observations , 1997, NIPS.

[7] Christopher M. Bishop,et al. GTM: The Generative Topographic Mapping , 1998, Neural Computation.

[8] Vin de Silva,et al. Graph approximations to geodesics on embedded manifolds , 2000 .

[9] J. Tenenbaum,et al. A global geometric framework for nonlinear dimensionality reduction. , 2000, Science.

[10] S T Roweis,et al. Nonlinear dimensionality reduction by locally linear embedding. , 2000, Science.

[11] Mikhail Belkin,et al. Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering , 2001, NIPS.

[12] David D. Denison,et al. Nonlinear estimation and classification , 2003 .