ABSTRACT
We study the problem of discovering a manifold that best preserves information relevant to a nonlinear regression. Solving this problem involves extending and uniting two threads of research. On the one hand, the literature on sufficient dimension reduction has focused on methods for finding the best linear subspace for nonlinear regression; we extend this to manifolds. On the other hand, the literature on manifold learning has focused on unsupervised dimensionality reduction; we extend this to the supervised setting. Our approach to solving the problem involves combining the machinery of kernel dimension reduction with Laplacian eigenmaps. Specifically, we optimize cross-covariance operators in kernel feature spaces that are induced by the normalized graph Laplacian. The result is a highly flexible method in which no strong assumptions are made on the regression function or on the distribution of the covariates. We illustrate our methodology on the analysis of global temperature data and image manifolds.
- Belkin, M., & Niyogi, P. (2003). Laplacian Eigenmaps for dimensionality reduction and data representation. Neural Computation, 15, 1373--1396. Google ScholarDigital Library
- Belkin, M., Niyogi, P., & Sindhwani, V. (2006). Manifold regularization: A geometric framework for learning from labeled and unlabeled examples. Journal of Machine Learning Research, 7, 2399--2434. Google ScholarDigital Library
- Bengio, Y., Paiement, J.-F., & Vincent, P. (2003). Out-of-sample extensions for LLE, Isomap, MDS, Eigenmaps, and spectral clustering (Technical Report 1238). Département d'Informatique et Recherche Opérationnelle, Université de Montréal.Google Scholar
- Chiaromonte, F., & Cook, R. D. (2002). Sufficient dimension reduction and graphics in regression. Annals of the Institute of Statistical Mathematics, 54, 768--795.Google ScholarCross Ref
- Coifman, R., Lafon, S., Lee, A., Maggioni, M., Nadler, B., Warner, F., & Zucker, S. (2005). Geometric diffusions as a tool for harmonic analysis and structure definition of data: Diffusion maps. Proceedings of the National Academy of Sciences USA, 102, 7426--7431.Google ScholarCross Ref
- Cook, R. D. (1998). Regression graphics. Wiley Inter-Science.Google Scholar
- Cook, R. D., & Li, B. (1991). Discussion of Li (1991). Journal of the American Statistical Association, 86, 328--332.Google Scholar
- Cook, R. D., & Yin, X. (2001). Dimension reduction and visualization in discriminant analysis (with discussion). Australian & New Zealand Journal of Statistics, 43, 147--199.Google ScholarCross Ref
- Donoho, D. L., & Grimes, C. (2003). Hessian eigenmaps: Locally linear embedding techniques for high-dimensional data. Proceedings of the National Academy of Sciences USA, 100, 5591--5596.Google ScholarCross Ref
- Fukumizu, K., Bach, F. R., & Jordan, M. I. (2004). Dimensionality reduction for supervised learning with reproducing kernel Hilbert spaces. Journal of Machine Learning Research, 5, 73--99. Google ScholarDigital Library
- Fukumizu, K., Bach, F. R., & Jordan, M. I. (2006). Kernel dimension reduction in regression (Technical Report). Department of Statistics, University of California, Berkeley.Google Scholar
- Globerson, A., & Tishby, N. (2003). Sufficient dimensionality reduction. Journal of Machine Learning Research, 3, 1307--1331. Google ScholarDigital Library
- Ham, J., Lee, D., Mika, S., & Schöölkopf, B. (2004). A kernel view of the dimensionality reduction of manifolds. Proceedings of the 21'st International Conference on Machine Learning. ACM. Google ScholarDigital Library
- Lafon, S. (2004). Diffusion maps and geometric harmonics. Doctoral dissertation, Yale University.Google Scholar
- Li, B., Zha, H., & Chiaramonte, F. (2005). Contour regression: A general approach to dimension reduction. The Annals of Statistics, 33, 1580--1616.Google ScholarCross Ref
- Li, K.-C. (1991). Sliced inverse regresion for dimension reduction. Journal of the American Statistical Association, 86, 316--327.Google ScholarCross Ref
- Li, K.-C. (1992). On principal Hessian directions for data visualization and dimension reduction: Another application of Stein's lemma. Journal of the American Statistical Association, 86, 316--342.Google ScholarCross Ref
- Remote Sensing Systems (2004). Microwave sounding units (MSU) data. sponsored by the NOAA Climate and Global Change Program. Data available at www.remss.com.Google Scholar
- Roweis, S., & Saul, L. (2000). Nonlinear dimensionality reduction by locally linear embedding. Science, 290, 2323--2326.Google ScholarCross Ref
- Sajama, & Orlitsky, A. (2005). Supervised dimensionality reduction using mixture models. Proceedings of the 22'nd International Conference on Machine Learning (pp. 760--767). ACM. Google ScholarDigital Library
- Sha, F., & Saul, L. (2005). Analysis and extension of spectral methods for nonlinear dimensionality reduction. Proceedings of the 22'nd International Conference on Machine Learning (pp. 785--792). ACM. Google ScholarDigital Library
- Tenenbaum, J., de Silva, V., & Langford, J. (2000). A global geometric framework for nonlinear dimensionality reduction. Science, 290, 2319--2322.Google ScholarCross Ref
- Yang, X., Fu, H., Zha, H., & Barlow, J. (2006). Semi-supervised nonlinear dimensionality reduction. Proceedings of the 23'rd International Conference on Machine Learning (pp. 1065--1072). ACM. Google ScholarDigital Library
- Regression on manifolds using kernel dimension reduction
Recommendations
Geometrically local embedding in manifolds for dimension reduction
In this paper, geometrically local embedding (GLE) is presented to discover the intrinsic structure of manifolds as a method in nonlinear dimension reduction. GLE is able to reveal the inner features of the input data in the lower dimension space while ...
Distance approximating dimension reduction of Riemannian manifolds
We study the problem of projecting high-dimensional tensor data on an unspecified Riemannian manifold onto some lower dimensional subspace without much distorting the pairwise geodesic distances between data points on the Riemannian manifold while ...
Max-Min Distance Analysis by Using Sequential SDP Relaxation for Dimension Reduction
We propose a new criterion for discriminative dimension reduction, max-min distance analysis (MMDA). Given a data set with C classes, represented by homoscedastic Gaussians, MMDA maximizes the minimum pairwise distance of these C classes in the selected ...
Comments