The Community for Technology Leaders
RSS Icon
Subscribe
Issue No.02 - March/April (2009 vol.15)
pp: 249-261
Stephen Ingram , University of British Columbia, Vancouver
Tamara Munzner , University of British Columbia, Vancouver
Marc Olano , University of Maryland Baltimore County, Baltimore
ABSTRACT
We present Glimmer, a new multilevel algorithm for multidimensional scaling designed to exploit modern graphics processing unit (GPU) hardware. We also present GPU-SF, a parallel, force-based subsystem used by Glimmer. Glimmer organizes input into a hierarchy of levels and recursively applies GPU-SF to combine and refine the levels. The multilevel nature of the algorithm makes local minima less likely while the GPU parallelism improves speed of computation. We propose a robust termination condition for GPU-SF based on a filtered approximation of the normalized stress function. We demonstrate the benefits of Glimmer in terms of speed, normalized stress, and visual quality against several previous algorithms for a range of synthetic and real benchmark datasets. We also show that the performance of Glimmer on GPUs is substantially faster than a CPU implementation of the same algorithm.
INDEX TERMS
Multivariate visualization, Information visualization
CITATION
Stephen Ingram, Tamara Munzner, Marc Olano, "Glimmer: Multilevel MDS on the GPU", IEEE Transactions on Visualization & Computer Graphics, vol.15, no. 2, pp. 249-261, March/April 2009, doi:10.1109/TVCG.2008.85
REFERENCES
[1] S. Arya, D.M. Mount, N.S. Netanyahu, R. Silverman, and A.Y. Wu, “An Optimal Algorithm for Approximate Nearest Neighbor Searching Fixed Dimensions,” J. ACM, vol. 45, no. 6, pp. 891-923, 1998.
[2] U. Brandes and C. Pich, “Eigensolver Methods for Progressive Multidimensional Scaling of Large Data,” Proc. 14th Int'l Symp. Graph Drawing (GD '06), pp. 42-53, 2006.
[3] M.M. Bronstein, A.M. Bronstein, R. Kimmel, and I. Yavneh, “Multigrid Multidimensional Scaling,” Numerical Linear Algebra with Applications, vol. 13, pp. 149-171, Mar.-Apr. 2006.
[4] M. Chalmers, “A Linear Iteration Time Layout Algorithm for Visualising High Dimensional Data,” Proc. Seventh IEEE Conf. Visualization, pp. 127-132, 1996.
[5] J. de Leeuw, “Applications of Convex Analysis to Multidimensional Scaling,” Recent Developments in Statistics, pp. 133-145, 1977.
[6] V. de Silva and J. Tenenbaum, “Sparse Multidimensional Scaling Using Landmark Points,” technical report, Stanford, 2004.
[7] C. Faloutsos and K.-I. Lin, “FastMap: A Fast Algorithm for Indexing, Data-Mining and Visualization of Traditional and Multimedia Datasets,” Proc. ACM SIGMOD '95, pp. 163-174, 1995.
[8] Y. Frishman and A. Tal, “Multi-Level Graph Layout on the GPU,” IEEE Trans. Visualization and Computer Graphics, vol. 13, no. 6, pp.1310-1319, Nov./Dec. 2007.
[9] Y. Frishman and A. Tal, “Online Dynamic Graph Drawing,” Proc. Eurographics/IEEE VGTC Symp. Visualization (EuroVis), 2007.
[10] E. Gansner, Y. Koren, and S. North, “Graph Drawing by Stress Majorization,” Proc. 12th Int'l Symp. Graph Drawing (GD '04), pp.239-250, 2004.
[11] F. Jourdan and G. Melancon, “Multiscale Hybrid MDS,” Proc. Eighth Int'l Conf. Information Visualization (IV '04), pp. 388-393, 2004.
[12] Y. Koren, “Graph Drawing by Subspace Optimization,” Proc. Eurographics/IEEE TCVG Symp. Visualization (VisSym '04), pp.65-74, 2004.
[13] Y. Koren, L. Carmel, and D. Harel, “ACE: A Fast Multiscale Eigenvectors Computation for Drawing Huge Graphs,” Proc. IEEE Symp. Information Visualization (InfoVis '02), p. 137, 2002.
[14] A. Krowne and M. Halbert, “An Initial Evaluation of Automated Organization for Digital Library Browsing,” Proc. Fifth ACM/IEEE-CS Joint Conf. Digital Libraries (JCDL '05), pp. 246-255, 2005.
[15] J.B. Kruskal, “Multidimensional Scaling by Optimizing Goodness of Fit to a Nonmetric Hypothesis,” Psychometrika, vol. 29, no. 1, pp.1-27, 1964.
[16] A. Morrison, G. Ross, and M. Chalmers, “Fast Multidimensional Scaling through Sampling, Springs and Interpolation,” Information Visualization, vol. 2, no. 1, pp. 68-77, 2003.
[17] nVidia, nVidia CUDA Compute Unified Device Architecture Programming Guide, Jan. 2007.
[18] J.D. Owens, D. Luebke, N. Govindaraju, M. Harris, J. Krüger, A.E. Lefohn, and T.J. Purcell, “A Survey of General-Purpose Computation on Graphics Hardware,” Computer Graphics Forum, vol. 26, no. 1, pp. 80-113, 2007.
[19] K. Perlin, “An Image Synthesizer,” Proc. ACM SIGGRAPH '85, pp.287-296, 1985.
[20] J. Platt, “FastMap, MetricMap, and Landmark MDS Are All Nyström Algorithms,” Proc. 10th Int'l Workshop Artificial Intelligence and Statistics, pp. 261-268, 2005.
[21] G. Reina and T. Ertl, “Implementing FastMap on the GPU: Considerations on General-Purpose Computation on Graphics Hardware,” Proc. Theory and Practice of Computer Graphics Conf. (TPCG '05), pp. 51-58, 2005.
[22] G. Ross and M. Chalmers, “A Visual Workspace for Constructing Hybrid Multidimensional Scaling Algorithms and Coordinating Multiple Views,” Information Visualization, vol. 2, no. 4, pp. 247-257, Dec. 2003.
[23] S.T. Roweis and L.K. Saul, “Nonlinear Dimensionality Reduction by Locally Linear Embedding,” Science, vol. 290, no. 5,500, Dec. 2000.
[24] W.S. Torgerson, “Multidimensional Scaling: I. Theory and Method,” Psychometrika, vol. 17, pp. 401-419, 1952.
[25] G. Young and A.S. Householder, “Discussion of a Set of Points in Terms of Their Mutual Distances,” Psychometrika, vol. 3, no. 1, Jan. 1938.
6 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool