Lisha Chen
Abstract
Multidimensional scaling (MDS) is the art of reconstructing pointsets (embeddings) from pairwise distance data, and as such it is at the basis of several approaches to nonlinear dimension reduction and manifold learning. At present, MDS lacks a unifying methodology as it consists of a discrete collection of proposals that differ in their optimization criteria, called "stress functions". To correct this situation we propose (1) to embed many of the extant stress functions in a parametric family of stress functions, and (2) to replace the ad hoc choice among discrete proposals with a principled parameter selection method. This methodology yields the following benefits and problem solutions: (a) It provides guidance in tailoring stress functions to a given data situation, responding to the fact that no single stress function dominates all others across all data situations; (b) the methodology enriches the supply of available stress functions; (c) it helps our understanding of stress functions by replacing the comparison of discrete proposals with a characterization of the effect of parameters on embeddings; (d) it builds a bridge to graph drawing, which is the related but not identical art of constructing embeddings from graphs.
- Ulas Akkucuk and J. Douglas Carroll. PARAMAP vs. Isomap: A comparison of two nonlinear mapping algorithms. Journal of Classification, 23(2):221-254, 2006.Google Scholar
- Ingwer Borg and Patrick J.F. Groenen. Modern Multidimensional Scaling: Theory and Applications. Springer-Verlag, 2005.Google Scholar
- Lisha Chen. Local Multidimensional Scaling for Nonlinear Dimension Reduction, Graph Layout and Proximity Analysis. PhD thesis, Ph.d. Thesis, University of Pennsylvania, Philadelphia, Pennsylvania, 2006.Google Scholar
- Lisha Chen and Andreas Buja. Local multidimensional scaling for nonlinear dimension reduction, graph drawing, and proximity analysis. Journal of the American Statistical Association, 104 (485):209-219, 2009.Google Scholar
- Ron Davidson and David Harel. Drawing graphs nicely using simulated annealing. ACM Transactions on Graphics (TOG), 15(4):301-331, 1996. Google Scholar
- Thomas M.J. Fruchterman and Edward M. Reingold. Graph drawing by force-directed placement. Software: Practice and Experience, 21(11):1129-1164, 1991. Google Scholar
- Emden Gansner, Yehuda Koren, and Stephen North. Graph drawing by stress majorization. In Graph Drawing, pages 239-250. Springer, 2005. Google Scholar
- Tomihisa Kamada and Satoru Kawai. An algorithm for drawing general undirected graphs. Information Processing Letters, 31(1):7-15, 1989. Google Scholar
- Yehuda Koren and Ali Çivril. The binary stress model for graph drawing. In Graph Drawing, pages 193-205. Springer, 2009. Google Scholar
- Joseph B. Kruskal. Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. Psychometrika, 29(1):1-27, 1964a.Google Scholar
- Joseph B. Kruskal. Nonmetric multidimensional scaling: A numerical method. Psychometrika, 29 (2):115-129, 1964b.Google Scholar
- Joseph B. Kruskal and Judith B. Seery. Designing network diagrams. In Proc. First General Conference on Social Graphics, pages 22-50, 1980.Google Scholar
- John A. Lee and Michel Verleysen. Rank-based quality assessment of nonlinear dimensionality reduction. In 16th European Symposium on Artificial Neural Networks (ESANN), Bruges, Belgium, pages 49-54, 2008.Google Scholar
- Fan Lu, Sündüz Keles, Stephen J Wright, and Grace Wahba. Framework for kernel regularization with application to protein clustering. Proceedings of the National Academy of Sciences of the United States of America, 102(35):12332-12337, 2005.Google Scholar
- Andreas Noack. Energy models for drawing clustered small world graphs. Technical report, Inst. of Computer Science, Brandenburg Technical University, Cottbus, Germany, 2003.Google Scholar
- Andreas Noack. Energy models for graph clustering. Journal of Graph Algorithms and Applications, 11(2):453-480, 2007.Google Scholar
- Andreas Noack. Modularity clustering is force-directed layout. Physical Review E, 79(2):026102, 2009.Google Scholar
- Sam T. Roweis and Lawrence K. Saul. Nonlinear dimensionality reduction by locally linear embedding. Science, 290(5500):2323-2326, 2000.Google Scholar
- John W. Sammon. A nonlinear mapping for data structure analysis. IEEE Transactions on Computers, 100(5):401-409, 1969. Google Scholar
- Bernhard Schölkopf, Alexander Smola, and Klaus-Robert Müller. Nonlinear component analysis as a kernel eigenvalue problem. Neural Computation, 10(5):1299-1319, 1998. Google Scholar
- Yoshio Takane, Forrest W. Young, and Jan De Leeuw. Nonmetric individual differences multidimensional scaling: An alternating least squares method with optimal scaling features. Psychometrika, 42(1):7-67, 1977.Google Scholar
- Joshua B. Tenenbaum, Vin De Silva, and John C. Langford. A global geometric framework for nonlinear dimensionality reduction. Science, 290(5500):2319-2323, 2000.Google Scholar
- Michael W. Trosset. Classical multidimensional scaling and laplacian eigenmaps. Presentation given at the 2006 Joint Statistical Meeting (Session 411), 2006.Google Scholar
- Jarkko Venna and Samuel Kaski. Local multidimensional scaling. Neural Networks, 19(6):889-899, 2006. Google Scholar
- Kilian Q. Weinberger, Fei Sha, Qihui Zhu, and Lawrence K. Saul. Graph Laplacian regularization for large-scale semidefinite programming. Advances in Neural Information Processing Systems, 19:1489, 2007.Google Scholar
Index Terms
Stress functions for nonlinear dimension reduction, proximity analysis, and graph drawing
Recommendations
Hybrid-mixed stress model for the nonlinear analysis of concrete structures
This communication reports on the nonlinear analysis of 2D concrete structures using an hybrid-mixed stress formulation and continuum damage mechanics. The finite element model adopted is based on the direct and independent approximation of three fields:...
Image analysis with nonlinear adaptive dimension reduction
ICIMCS '11: Proceedings of the Third International Conference on Internet Multimedia Computing and ServiceIn multimedia applications, dimension reduction is essential to the effectiveness and efficiency of an algorithm due to the curse of dimensionality. Recently, its adaptive variants have received considerable attention in unsupervised learning since a ...
Free-vibration analysis of arches based on the hybrid-mixed formulation with consistent quadratic stress functions
The well-known two-node hybrid-mixed element with linear displacement interpolation functions and constant stress resultant functions proposed by Saleeb yields locking-free results in static analysis but shows unsatisfactory results in vibration ...
Comments