Spectral Sequencing Based on Graph Distance Rong Liu, Hao Zhang, Oliver van Kaick {lrong, haoz, cs.sfu.ca {lrong, haoz, cs.sfu.ca.

Slides:



Advertisements
Similar presentations
Pattern Recognition and Machine Learning
Advertisements

Coherent Laplacian 3D protrusion segmentation Oxford Brookes Vision Group Queen Mary, University of London, 11/12/2009 Fabio Cuzzolin.
Alignment Visual Recognition “Straighten your paths” Isaiah.
Partial Differential Equations
Cluster Analysis: Basic Concepts and Algorithms
Principal Component Analysis Based on L1-Norm Maximization Nojun Kwak IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.
Nonlinear Dimension Reduction Presenter: Xingwei Yang The powerpoint is organized from: 1.Ronald R. Coifman et al. (Yale University) 2. Jieping Ye, (Arizona.
1 Appendix B: Solving TSP by Dynamic Programming Course: Algorithm Design and Analysis.
« هو اللطیف » By : Atefe Malek. khatabi Spring 90.
1er. Escuela Red ProTIC - Tandil, de Abril, 2006 Principal component analysis (PCA) is a technique that is useful for the compression and classification.
Graph Laplacian Regularization for Large-Scale Semidefinite Programming Kilian Weinberger et al. NIPS 2006 presented by Aggeliki Tsoli.
Watermarking 3D Polygonal Meshes in the Mesh Spectral Domain Ryutarou Ohbuchi, Shigeo Takahashi, Takahiko Miyazawa, Akio Mukaiyama.
10/11/2001Random walks and spectral segmentation1 CSE 291 Fall 2001 Marina Meila and Jianbo Shi: Learning Segmentation by Random Walks/A Random Walks View.
Chapter 3 The Greedy Method 3.
Complexity 16-1 Complexity Andrei Bulatov Non-Approximability.
6/26/2006CGI'06, Hangzhou China1 Sub-sampling for Efficient Spectral Mesh Processing Rong Liu, Varun Jain and Hao Zhang GrUVi lab, Simon Fraser University,
Principal Component Analysis CMPUT 466/551 Nilanjan Ray.
Principal Component Analysis
One-Shot Multi-Set Non-rigid Feature-Spatial Matching
ETHEM ALPAYDIN © The MIT Press, Lecture Slides for.
Branch and Bound Searching Strategies
Principal Component Analysis
L16: Micro-array analysis Dimension reduction Unsupervised clustering.
1 Numerical geometry of non-rigid shapes Spectral Methods Tutorial. Spectral Methods Tutorial 6 © Maks Ovsjanikov tosca.cs.technion.ac.il/book Numerical.
Computer Algorithms Integer Programming ECE 665 Professor Maciej Ciesielski By DFG.
Basic Concepts and Definitions Vector and Function Space. A finite or an infinite dimensional linear vector/function space described with set of non-unique.
Three Algorithms for Nonlinear Dimensionality Reduction Haixuan Yang Group Meeting Jan. 011, 2005.
Clustering In Large Graphs And Matrices Petros Drineas, Alan Frieze, Ravi Kannan, Santosh Vempala, V. Vinay Presented by Eric Anderson.
Preference Analysis Joachim Giesen and Eva Schuberth May 24, 2006.
1 Branch and Bound Searching Strategies 2 Branch-and-bound strategy 2 mechanisms: A mechanism to generate branches A mechanism to generate a bound so.
A Global Geometric Framework for Nonlinear Dimensionality Reduction Joshua B. Tenenbaum, Vin de Silva, John C. Langford Presented by Napat Triroj.
Nonlinear Dimensionality Reduction Approaches. Dimensionality Reduction The goal: The meaningful low-dimensional structures hidden in their high-dimensional.
1 Shortest Path Calculations in Graphs Prof. S. M. Lee Department of Computer Science.
Graph-based consensus clustering for class discovery from gene expression data Zhiwen Yum, Hau-San Wong and Hongqiang Wang Bioinformatics, 2007.
Domain decomposition in parallel computing Ashok Srinivasan Florida State University COT 5410 – Spring 2004.
Manifold learning: Locally Linear Embedding Jieping Ye Department of Computer Science and Engineering Arizona State University
Overview of Kernel Methods Prof. Bennett Math Model of Learning and Discovery 2/27/05 Based on Chapter 2 of Shawe-Taylor and Cristianini.
Gene expression & Clustering (Chapter 10)
Summarized by Soo-Jin Kim
Linear Least Squares Approximation. 2 Definition (point set case) Given a point set x 1, x 2, …, x n  R d, linear least squares fitting amounts to find.
The greedy method Suppose that a problem can be solved by a sequence of decisions. The greedy method has that each decision is locally optimal. These.
Introduction to variable selection I Qi Yu. 2 Problems due to poor variable selection: Input dimension is too large; the curse of dimensionality problem.
Graph Algorithms. Definitions and Representation An undirected graph G is a pair (V,E), where V is a finite set of points called vertices and E is a finite.
IEEE TRANSSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 3: LINEAR MODELS FOR REGRESSION.
Computer Vision Lab. SNU Young Ki Baik Nonlinear Dimensionality Reduction Approach (ISOMAP, LLE)
Spectral Analysis based on the Adjacency Matrix of Network Data Leting Wu Fall 2009.
Semantic Wordfication of Document Collections Presenter: Yingyu Wu.
A Convergent Solution to Tensor Subspace Learning.
Copyright © 2007 Pearson Education, Inc. Publishing as Pearson Addison-Wesley. Ver Chapter 13: Graphs Data Abstraction & Problem Solving with C++
Domain decomposition in parallel computing Ashok Srinivasan Florida State University.
1 Microarray Clustering. 2 Outline Microarrays Hierarchical Clustering K-Means Clustering Corrupted Cliques Problem CAST Clustering Algorithm.
Branch and Bound Searching Strategies
Chapter 13 Discrete Image Transforms
1 Objective To provide background material in support of topics in Digital Image Processing that are based on matrices and/or vectors. Review Matrices.
Spectral Methods for Dimensionality
Spectral partitioning works: Planar graphs and finite element meshes
Principal Component Analysis (PCA)
Spectral Methods Tutorial 6 1 © Maks Ovsjanikov
Computability and Complexity
Outline Nonlinear Dimension Reduction Brief introduction Isomap LLE
Spectral Clustering Eric Xing Lecture 8, August 13, 2010
3. Brute Force Selection sort Brute-Force string matching
Feature space tansformation methods
Generally Discriminant Analysis
3. Brute Force Selection sort Brute-Force string matching
Principal Component Analysis
Chapter 14 Graphs © 2011 Pearson Addison-Wesley. All rights reserved.
Clustering.
3. Brute Force Selection sort Brute-Force string matching
Presentation transcript:

Spectral Sequencing Based on Graph Distance Rong Liu, Hao Zhang, Oliver van Kaick {lrong, haoz, cs.sfu.ca {lrong, haoz, cs.sfu.ca Introduction School of Computing Science, Simon Fraser University, Burnaby, Canada. Result Spectral Sequencing Practical Issues Linear Graph Layout Problem Our Approach Problem Spectral–Embed Graph Vertices in High Dimensional Space - (step 1) Project the Embedding to 1-D to obtain the Sequence - (step 2) Sub-sample to Reduce Complexity Avoid the Centering of Kernel Matrix K Comparison of Different Algorithms Models a e b c d f g h  We are interested in generating optimal linear mesh layouts (a sequence of mesh primitives) according to certain cost functions. Optimal mesh layouts generation has various applications: one particular example is Streaming Meshes [Isenburg and Lindstrom, 2005], where streamability is measured using span and width.  We cast optimal mesh layout generation problem as a more general graph layout generation problem.  Consider an edge-weighted graph G =( V, E, w ), a layout of G is a labeling  : V→{1, 2, …, n}.  To define an optimal labeling , different cost functions C (, w ) can be considered, such as:  span:, is the largest index difference of the two end vertices of an edge in the labeling. One typical algorithm to generate labeling with small span is Cuthill-Mckee (CM) scheme.  width:, measures, at a certain point of the linear layout, the number of edges for which only one end vertex has been encountered. Fiedler vector, the eigenvector corresponding to the second smallest eigenvalue of the graph Laplacian, is a frequently used heuristic to derive labeling with small width.  Although different cost functions can be considered for optimal graph layouts, we observe that it is generally desirable to have close vertices in the graph to be close in the layout. For this purpose, our algorithm (step 1) embeds graph vertices in a high dimensional space while keeping their relative distances in the graph. Then (step 2) the embedded vertices are projected onto a vector to generate the layout, with the least distance distortion.  Given a weighted graph G =( V, E, w ), build matrix which encodes the pair- wise distances between graph vertices.  Convert D to the affinity matrix K, where K ij =exp(- D ij / 2   ).  Center K by setting.  Consider K as a kernel matrix. With the concept of kernel method where K ij = ‹( v i ), ( v j )›, a vertex v is mapped to ( v ) in a high dimensional feature space. Thus K implicitly defines an embedding ( V ) = {( v ) | v V }. The order of distances among graph vertices is preserved: closer vertices in the graph (with smaller D ij ) are also closer in the feature space (with smaller  ij ). Order of Distances among Vertices are Preserved in the Embedding  The squared distance between two points u and v in the feature space is:  ij = ||( v i ), ( v j )|| 2 = ( v i ) T ( v i ) + ( v j ) T ( v j ) - 2 ( v i ) T ( v j ) = exp(- D ij / 2   ).  For the sake of simplicity and efficiency, the embedding ( V ) is simply projected onto a vector p* along which the pair-wise distances are mostly preserved. The projections are sorted to give the sequence.  The projection vector is chosen as. The underlying reason is simple. Since the length of the projection of any vector onto p is smaller than its original length, the objective function above tends to preserve the collective lengths of all the projections on p*. As the result, the relative distances between vertices are mostly preserved, in abovementioned sense, after the projection.  It can be proven that the sequence obtained by ordering the projections of ( V ) on p* is identical to that obtained by ordering the elements of the largest eigenvector U 1 of K. Therefore we simply compute U 1 and order its elements to obtain the sequence. The spectral sequencing algorithm is not practical yet due to the high computational overhead involved in the construction and the eigenvalue decomposition of K. We solve this problem by resorting to sub-sampling.  To build K, the overhead comes from the construction of D where Dijkstra shortest path algorithm is applied to all vertices. With sub-sampling, we only compute a small subset of rows of K, reducing the complexity from O ( n 2 log( n )) to O ( ln log( n )), where n is the problem size and l « n is the number of samples.  To compute the eigenvectors of K when only its partial rows are available, we use Nyström method. Without loss of generality, consider the sampled rows, [ A B ], are the leading rows, where A encodes the affinities between samples and B contains the cross-affinities between samples and the remaining. Nyström method computes the approximate eigenvectors in columns of as follows: partial kernel matrix: eigenvalue decomposition of A : approximate leading eigenvectors of K  As mentioned before, the kernel matrix K needs to be centered before the eigenvalue decomposition. When sub-sampling is applied, the problem is that K cannot be centered by only knowing the sampled rows, [ A B ].  Denote by the centered kernel matrix, it can be formally shown that when the kernel width  is sufficiently large, its largest eigenvector is close to the second largest eigenvector of un-centered K. Resultantly, we do not center K explicitly; instead, we simply use its second largest eigenvector for sequencing. To verify its effectiveness, we take a graph with 600 vertices and construct its corresponding K and. Then we compute the L2 norm of the difference between, the largest eigenvector of and, and, the second largest eigenvector of K, against the size of As we can see from the plot, the difference between the two eigenvectors is negligible when  is sufficiently large. In our application, we set  to the average distance computed. Conclusion and Future Work  The results plotted are obtained on the mesh primal graph. Our approach outperforms Laplacian scheme in terms of span and CM scheme in terms of width. Therefore in applications where both measures are important, our approach could potentially provide a better trade-off.  In the future, we plan to investigate the influence of kernel functions, which is used to convert distance matrix D to affinity matrix K, on the distribution of points in the feature space, so as to improve the performance of our approach for both span and width, and also other cost functions. It is also interesting to study other techniques to obtain the sequence (layout) from the high dimensional embedding.  We compare our approach with CM and Laplacian schemes, which are effective for minimizing span and width, respectively. Another four cost functions are also considered.