Abstract

Nonlinear dimensionality reduction is formulated here as the problem of trying to find a Euclidean feature-space embedding of a set of observations that preserves as closely as possible their intrinsic metric structure – the distances between points on the observation manifold as measured along geodesic paths. Our isometric feature mapping procedure, or isomap, is able to reliably recover low-dimensional nonlinear structure in realistic perceptual data sets, such as a manifold of face images, where conventional global mapping methods find only local minima. The recovered map provides a canonical set of globally meaningful features, which allows perceptual transformations such as interpolation, extrapolation, and analogy – highly nonlinear transformations in the original observation space – to be computed with simple linear operations in feature space. 1

Citations

2715	Neural networks: A comprehensive foundation (2nd ed.). Upper Saddle River - Haykin - 1999
1158	Self-organization and associative memory - Kohonen - 1984
570	Networks for approximation and learning - Poggio, Girosi - 1990
320	Multidimensional Scaling - Cox, Cox - 1994
258	Designing and building parallel programs - Foster - 1995
157	Topology representing networks - Martinetz, Schulten - 1994
100	Nonlinear dimensionality reduction - DeMers, Cottrell - 1993
52	Recognizing handwritten digits using mixtures of linear models - Hinton, Revow, et al. - 1995
48	I.: GTM: The Generative Topographic Mapping. Neural Computation - Bishop, Svensen, et al. - 1998
48	Nonlinear image interpolation using manifold learning - Bregler, Omohundro
2	A Variational Principle for Modelbased Morphing, NIPS-9 - Saul, Jordan - 1997
1	Representations for visual learning, Science 272 - Beymer - 1995