Diffusion Geometries, and multiscale Harmonic Analysis on graphs and complex data sets. Multiscale diffusion geometries, “Ontologies and knowledge building”

Slides:



Advertisements
Similar presentations
Applications of one-class classification
Advertisements

CHAPTER 13: Alpaydin: Kernel Machines
Aggregating local image descriptors into compact codes
PARTITIONAL CLUSTERING
A Geometric Perspective on Machine Learning 何晓飞 浙江大学计算机学院 1.
Salvatore giorgi Ece 8110 machine learning 5/12/2014
Input Space versus Feature Space in Kernel- Based Methods Scholkopf, Mika, Burges, Knirsch, Muller, Ratsch, Smola presented by: Joe Drish Department of.
Nonlinear Dimension Reduction Presenter: Xingwei Yang The powerpoint is organized from: 1.Ronald R. Coifman et al. (Yale University) 2. Jieping Ye, (Arizona.
3D Shape Histograms for Similarity Search and Classification in Spatial Databases. Mihael Ankerst,Gabi Kastenmuller, Hans-Peter-Kriegel,Thomas Seidl Univ.
Presented by: Mingyuan Zhou Duke University, ECE April 3, 2009
Non-linear Dimensionality Reduction CMPUT 466/551 Nilanjan Ray Prepared on materials from the book Non-linear dimensionality reduction By Lee and Verleysen,
. Markov Chains as a Learning Tool. 2 Weather: raining today40% rain tomorrow 60% no rain tomorrow not raining today20% rain tomorrow 80% no rain tomorrow.
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.
Ronald R. Coifman , Stéphane Lafon, 2006
Principal Component Analysis CMPUT 466/551 Nilanjan Ray.
Region Segmentation. Find sets of pixels, such that All pixels in region i satisfy some constraint of similarity.
Prénom Nom Document Analysis: Data Analysis and Clustering Prof. Rolf Ingold, University of Fribourg Master course, spring semester 2008.
Three Algorithms for Nonlinear Dimensionality Reduction Haixuan Yang Group Meeting Jan. 011, 2005.
USER VERIFICATION SYSTEM. Scope Web Interface RGB separation Pervasive.
A Global Geometric Framework for Nonlinear Dimensionality Reduction Joshua B. Tenenbaum, Vin de Silva, John C. Langford Presented by Napat Triroj.
Smart Traveller with Visual Translator for OCR and Face Recognition LYU0203 FYP.
Pattern Recognition. Introduction. Definitions.. Recognition process. Recognition process relates input signal to the stored concepts about the object.
Atul Singh Junior Undergraduate CSE, IIT Kanpur.  Dimension reduction is a technique which is used to represent a high dimensional data in a more compact.
Nonlinear Dimensionality Reduction by Locally Linear Embedding Sam T. Roweis and Lawrence K. Saul Reference: "Nonlinear dimensionality reduction by locally.
Diffusion Maps and Spectral Clustering
Clustering Unsupervised learning Generating “classes”
Nonlinear Dimensionality Reduction Approaches. Dimensionality Reduction The goal: The meaningful low-dimensional structures hidden in their high-dimensional.
Dimensionality Reduction: Principal Components Analysis Optional Reading: Smith, A Tutorial on Principal Components Analysis (linked to class webpage)
MMDS- Stanford 2008 Harmonic Analysis, diffusion geometries and Multi Scale organizations of data and matrices. R.R Coifman Department of Mathematics,Yale.
BACKGROUND LEARNING AND LETTER DETECTION USING TEXTURE WITH PRINCIPAL COMPONENT ANALYSIS (PCA) CIS 601 PROJECT SUMIT BASU FALL 2004.
Data Reduction. 1.Overview 2.The Curse of Dimensionality 3.Data Sampling 4.Binning and Reduction of Cardinality.
Texture scale and image segmentation using wavelet filters Stability of the features Through the study of stability of the eigenvectors and the eigenvalues.
Line detection Assume there is a binary image, we use F(ά,X)=0 as the parametric equation of a curve with a vector of parameters ά=[α 1, …, α m ] and X=[x.
Computer Vision Lab. SNU Young Ki Baik Nonlinear Dimensionality Reduction Approach (ISOMAP, LLE)
Diffusion Geometries in Document Spaces. Multiscale Harmonic Analysis. R.R. Coifman, S. Lafon, A. Lee, M. Maggioni, B.Nadler. F. Warner, S. Zucker. Mathematics.
Handwritten digit recognition
CVPR2013 Poster Detecting and Naming Actors in Movies using Generative Appearance Models.
Manifold learning: MDS and Isomap
Nonlinear Dimensionality Reduction Approach (ISOMAP)
Tony Jebara, Columbia University Advanced Machine Learning & Perception Instructor: Tony Jebara.
CS 8751 ML & KDDData Clustering1 Clustering Unsupervised learning Generating “classes” Distance/similarity measures Agglomerative methods Divisive methods.
CSE 446 Dimensionality Reduction and PCA Winter 2012 Slides adapted from Carlos Guestrin & Luke Zettlemoyer.
Math 285 Project Diffusion Maps Xiaoyan Chong Department of Mathematics and Statistics San Jose State University.
 In the previews parts we have seen some kind of segmentation method.  In this lecture we will see graph cut, which is a another segmentation method.
Ultra-high dimensional feature selection Yun Li
Link Analysis Algorithms Page Rank Slides from Stanford CS345, slightly modified.
Stereo March 8, 2007 Suggested Reading: Horn Chapter 13.
Geometric diffusions as a tool for harmonic analysis and structure definition of data By R. R. Coifman et al. The second-round discussion* on * The first-round.
CSCI 631 – Foundations of Computer Vision March 15, 2016 Ashwini Imran Image Stitching.
Cluster Analysis What is Cluster Analysis? Types of Data in Cluster Analysis A Categorization of Major Clustering Methods Partitioning Methods.
1 Modification of Correlation Kernels in SVM, KPCA and KCCA in Texture Classification Yo Horikawa Kagawa University, Japan.
Spectral Methods for Dimensionality
3.1 Clustering Finding a good clustering of the points is a fundamental issue in computing a representative simplicial complex. Mapper does not place any.
Intrinsic Data Geometry from a Training Set
We propose a method which can be used to reduce high dimensional data sets into simplicial complexes with far fewer points which can capture topological.
Data Mining, Neural Network and Genetic Programming
9.3 Filtered delay embeddings
Unsupervised Riemannian Clustering of Probability Density Functions
Spectral Methods Tutorial 6 1 © Maks Ovsjanikov
3.1 Clustering Finding a good clustering of the points is a fundamental issue in computing a representative simplicial complex. Mapper does not place any.
Jianping Fan Dept of CS UNC-Charlotte
Outline Nonlinear Dimension Reduction Brief introduction Isomap LLE
Learning with information of features
CV: Matching in 2D Matching 2D images to 2D images; Matching 2D images to 2D maps or 2D models; Matching 2D maps to 2D maps MSU CSE 803 Stockman.
Principal Component Analysis (PCA)
Principal Component Analysis
Chap 8. Instance Based Learning
Lecture 15: Least Square Regression Metric Embeddings
Machine Learning – a Probabilistic Perspective
The “Margaret Thatcher Illusion”, by Peter Thompson
Presentation transcript:

Diffusion Geometries, and multiscale Harmonic Analysis on graphs and complex data sets. Multiscale diffusion geometries, “Ontologies and knowledge building” Ronald Coifman Applied Mathematics Yale university.

Conventional nearest neighbor search, compared with a diffusion search. The data is a pathology slide,each pixel is a digital document (spectrum below for each class )

One of our goals is to report on mathematical tools used in machine learning, document and web browsing, bio informatics, and many other data mining activities. The remarkable observation is that basic Geometric Harmonic Analysis of empirical Markov processes provides a unified mathematical structure which encapsulates most successful methods in these areas. These methods enable global descriptions of objects verifying microscopic relations (like calculus). We relate these ideas to methods of classical Harmonic analysis, like Calderon Zygmund theory in which Fourier analysis and multiscale geometry merge.

This simple point is illustrated below Each puzzle piece is linked to its neighbors ( in feature space ) the network of links forms a sphere. A parametrization of the sphere can be obtained from the eigenvectors of the inference relation (diffusion operator)

A simple empirical diffusion matrix A can be constructed as follows Let represent normalized data,we “soft truncate” the covariance matrix as A is a renormalized Markov version of this matrix The eigenvectors of this matrix provide a local non linear principal component analysis of the data. Whose entries are the diffusion coordinates These are also the eigenfunctions of the discrete Graph Laplace Operator. This map is a diffusion (at time t) embedding into Euclidean space

The First two eigenfunctions organize the small images which were provided in random order, in fact assembling the 3D puzzle.

A two dimensional map created by the Diffusion Map algorithm for 400 MMPI-2 examinees. The distance between two people was measured as the difference between their responses. The color corresponds to the score each examinee received on the depression scale. New subjects need to be placed in this tabulation of responders.

The following image indicates that graphs may have clusters at different scales.

A very simple way to build a hierarchical multiscale structure is as follows. We define the diffusion distance between two subsets E and F as : Start by considering small disjoint clusters of nearest neighbors. Form a graph of these clusters where the distance is defined with t=1. Repeat on the graph of these clusters doubling the time, etc

4 Gaussian Clouds

A simple application of signal processing on data,or data filters is Feature based diffusion algorithms. A simple application of signal processing on data,or data filters is Feature based diffusion algorithms. Given an image, associate with each pixel p a vector v(p) of features. For example a spectrum, or the 5x5 subimage centered at the pixel,or any combination of features. Define a Markov filter as The various powers of A or polynomials in A provide filters which account for feature similarity between pixels.

Feature diffusion filtering of the noisy Lenna image is achieved by associating with each pixel a feature vector (say the 5x5 subimage centerd at the pixel) this defines a Markov diffusion matrix which is used to filter the image,as was done in for the spiral in the preceding slide

The data is given as a random cloud, the filter organizes the data. The colors are not part of the data