Geometric Inference for Probability distributions

Size: px
Start display at page:

Download "Geometric Inference for Probability distributions"

Transcription

1 Geometric Inference for Probability distributions F. Chazal 1 D. Cohen-Steiner 2 Q. Mérigot 2 1 Geometrica, INRIA Saclay, 2 Geometrica, INRIA Sophia-Antipolis 2009 July 8 F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 1 / 32

2 Outline 1 Geometric inference for measures. 2 Distance to a probability measures. 3 Applications F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 2 / 32

3 Motivation What is the (relevant) topology/geometry of a point cloud data set in R d? Motivations : Reconstruction, manifold learning and NLDR, clustering and segmentation, etc.... Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 3 / 32

4 Geometric Inference Question Given an approximation C of a geometric object K, what geometric and topological quantities of K is it possible to approximate, knowing only C? The answer depends on the considered class of objects and a notion of distance between the objects (approximation) ; Some positive answers for a large class of compact sets endowed with the Hausdor distance. In this talk : - can the considered objects be probability measures on R d? - motivation : allowing approximations to have outliers or to be corrupted by non local noise. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 4 / 32

5 Distance functions for geometric inference Distance function and Hausfor distance Distance to a compact K R d : d K : x inf p K x p Hausdor distance between compact sets K, K R d : d H (K, K ) = inf x R d d K (x) d K (x) Replace K and C by d K and d C. Compare the topology of the osets K r = d 1 K ([0, r]) and C r = d 1 ([0, r]). C F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 5 / 32

6 Stability properties of the osets Topological/geometric properties of the osets of K are stable with respect to Hausdor approximation : 1. Topological stability of the osets of K (CCSL'06, NSW'06). 2. Approximate normal cones (CCSL'08). 3. Boundary measures (CCSM'07), curvature measures (CCSLT'09), Voronoi covariance measures (GMO'09). F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 6 / 32

7 The problem of outliers If K = K {x} where d K (x) > R, then d K d K > R : oset-based inference methods fail! Question : Can we generalized the previous approach by replacing the distance function by a distance-like function having a better behavior with respect to noise and outliers? F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 7 / 32

8 The three main ingredients for stability The stability in distance-based geometric inference relies on the three following facts : 1 the 1-Lipschitz property for d K ; 2 the 1-concavity on the function d 2 K : x x 2 d 2 K (x) is convex. 3 the stability of the map K d K : d K d K = sup x R d d K (x) d K (x) = dh(k, K ) A map φ : R d R which veries (1) and (2) is called distance-like. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 8 / 32

9 Replacing compact sets by measures A measure µ is a mass distribution on R d. Mathematically, it is dened as a map µ that takes a (Borel) subset B R d and outputs a nonnegative number µ(b). Moreover we ask that if (B i ) are disjoint subsets, µ ( i N B ) i = i N µ(b i) µ(b) corresponds to to the mass of µ contained in B a point cloud C = {p 1,..., p n } denes a measure µ C = 1 n i δ p i the volume form on a k-dimensional submanifold M of R d denes a measure vol k M. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 9 / 32

10 Distance between measures The Wasserstein distance dw(µ, ν) between two probability measures µ, ν quanties the optimal cost of pushing µ onto ν, the cost of moving a small mass dx from x to y being x y 2 dx. c 1 d 1 c i π i,j d j µ ν π 1 µ and ν are discrete measures : µ = i c iδ xi, ν = j d jδ yj with j d j = i c i. 2 Transport plan : set of coecients π ij 0 with i π ij = d j and j π ij = c i. 3 Cost of a transport plan ( ) 1/2 C(π) = ij x i y j 2 π ij 4 dw(µ, ν) := inf π C(π) F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 10 / 32

11 Wasserstein distance Examples : 1 If #C 1 = #C 2, then d 2 W (µ C 1, µ C2 ) is the cost of a minimal least-square matching between C 1 and C 2 ; 2 If C = {p 1,..., p n } and C = {p 1,..., p n k 1, o 1,..., o k } with d(o i, C) = R, then dh(c, C ) R while dw(µ C, µ C ) m(r + diam(c)) with m = k n ; 3 If µ is a probability measure, dw(µ N (0, σ), µ) σ ; 4 If X 1,..., X N are iid with law µ, then (in general), 1 N N δ i=1 X i converges to µ whp as N tends to. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 11 / 32

12 Outline 1 Geometric inference for measures. 2 Distance to a probability measures. 3 Applications F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 12 / 32

13 The distance to a measure Distance function to a measure, rst attempt Let m ]0, 1[ be a positive mass, and µ a probability measure on R d : δ µ,m (x) = inf {r > 0; µ(b(x, r)) > m} x δ µ,m(x) µ(b(x, δ µ,m(x)) = m µ δ µ,m is the smallest distance needed to attain a mass of at least m ; Coincides with the distance to the k-th neighbor when m = k/n and µ = 1 n n i=1 δ p i : δ µ,k/n (µ) = x p k C (x). F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 13 / 32

14 Unstability of µ δ µ,m Distance to a measure, rst attempt Let m ]0, 1[ be a positive mass, and µ a probability measure on R d : δ µ,m (x) = inf {r > 0; µ(b(x, r)) > m} Unstability under Wasserstein perturbations : µ ε = (1/2 ε)δ 0 + (1/2 + ε)δ 1 for ε > 0 : x < 0, δ µε,1/2(x) = x 1 for ε = 0 : x < 0, δ µ0,1/2(x) = x 0 Consequence : the map µ δ µ,m C 0 (R d ) is discontinuous whatever the (reasonable) topology on C 0 (R d ). F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 14 / 32

15 The distance function to a measure. Denition If µ is a measure on R d and m 0 > 0, one let : ( 1 d µ,m0 : x R d m0 ) 1/2 δ 2 m µ,m(x)dm 0 0 Example. Let C = {p 1,..., p n } and µ = 1 n n δ i=1 p i. Let p k C (x) denote the kth nearest neighbor to x in C, and set m 0 = k 0 /n : ( 1 d µ,m0 (x) = k 0 k 0 k=1 x p k C (x) 2 ) 1/2 F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 15 / 32

16 The distance function to a discrete measure. Example (continued) Let C = {p 1,..., p n } and µ = 1 n n i=1 δ p i. Let p k C (x) denote the kth nearest neighbor to x in C, and set m 0 = k 0 /n : ( 1 d µ,m0 (x) = d µ,m0 (x) = k 0 k 0 k=1 x p k C (x) 2 ) 1/2 1 k0 [ k 0 k=1 x p k C (x) ] d µ,m0 (x) F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 16 / 32

17 Another expression for d µ,m0 Proposition The distance d µ,m0 (x) coincides with the partial Wasserstein distance between the Dirac mass m 0 δ x and µ. More precisely : m0 d µ,m0 (x) = min{dw(m 0 δ x, ν); ν µ and mass(ν) = m 0 } { ( ) } 1/2 = min y x 2 dν(y) ; ν µ, mass(ν) = m 0 R d Let µ x,m0 be a measure realizing this minimum. The measure µ x,m0 gives mass to the multiple projections of x on µ ; For the point cloud case, when m 0 = k 0 /n and x is not on a k-voronoï face, µ x,m0 = k 0 k=1 1 n δ p k C (x) F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 17 / 32

18 1-Concavity of the squared distance function Regularity m 0 d 2 µ,m 0 (x + h) = x + h z d dµ x+h,m0 (z) R d x + h z d dµ x,m0 (z) R d m 0 d 2 µ,m 0 (x) + 2 h x z dµ x,m0 (z) + m 0 h 2 R d That is : d 2 µ,m 0 (x + h) d 2 µ,m 0 (x) + h d 2 µ,m 0 (x) + h 2 with d 2 µ,m 0 (x) := 2m 1 0 (x z)dµ x,m0 (z) R d F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 18 / 32

19 Theorem The distance function d µ,m0 is distance-like,ie. 1 the function x d µ,m0 (x) is 1-Lipschitz ; 2 the function x x 2 d 2 µ,m 0 (x) is convex ; Theorem The map µ d µ,m0 1 m0 Lipschitz, ie from probability measures to continuous functions is dµ,m0 d µ,m 0 1 m0 dw(µ, µ ) F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 19 / 32

20 Outline 1 Geometric inference for measures. 2 Distance to a probability measures. 3 Applications F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 20 / 32

21 Consequences of the previous properties 1 existence of an analogous to the medial axis 2 stability of a ltered version of it (as with the µ-medial axis) under Wasserstein perturbation 3 stability of the critical function of a measure 4 the gradient d µ,m0 is L 1 -stable 5... = the distance functions d µ,m0 share many stability and regularity properties with the usual distance function. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 21 / 32

22 Example : square with outliers 10% outliers, k = 150 δ µ,m0, m 0 = 1/10. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 22 / 32

23 Example : square with outliers d µ,m0 d µ,m0 F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 23 / 32

24 A 3D example Reconstruction of an oset from a noisy dataset, with 10% outliers. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 24 / 32

25 A reconstruction theorem Theorem Let µ be a probability measure of dimension at most k > 0 with compact support K R d such that r α (K) > 0 for some α (0, 1]. For any 0 < η < r α (K), there exists positive constants m 1 = m 1 (µ, α, η) > 0 and C = C(m 1 ) > 0 such that : for any m 0 < m 1 and any probability measure µ such that W 2 (µ, µ ) < C m 0, the sublevel set d 1 µ,m 0 ((, η]) is homotopy equivalent (and even isotopic) to the osets d 1 K ([0, r]) of K for 0 < r < r α (K). F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 25 / 32

26 k-nn density estimation vs distance to a measure data Density is estimated using x m 0 ω d 1 (δ µ,m 0 (x)), m 0 = 150/1200. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 26 / 32

27 k-nn density estimation vs distance to a measure density log(1 + 1/d µ,m0 ) Density is estimated using x (Devroye-Wagner '77). m 0 vol d (B(x,δ µ,m 0 (x))), m 0 = 150/1200. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 27 / 32

28 k-nn density estimation vs distance to a measure 1 the gradient of the estimated density can behave wildly 2 exhibits peaks near very dense zone 1. can be xed using d µ,m0 (because of the semiconcavity) 2. shows that the distance function is a better-behaved geometric object to associate to a measure. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 28 / 32

29 k-nn density estimation vs distance to a measure 1 the gradient of the estimated density can behave wildly 2 exhibits peaks near very dense zone 1. can be xed using d µ,m0 (because of the semiconcavity) 2. shows that the distance function is a better-behaved geometric object to associate to a measure. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 29 / 32

30 Pushing data along the gradient of d µ,m0 Mean-Shift like algorithm (Comaniciu-Meer '02) Theoretical guarantees on the convergence of the algorithm and smoothness of trajectories.. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 30 / 32

31 Pushing data along the gradient of d µ,m0. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 31 / 32

32 Summary µ d µ,m0 provide a way to associate geometry to a measure in Euclidean space. d µ,m0 is robust to Wasserstein perturbations : outliers and noise are easily handled (no assumption on the nature of the noise). d µ,m0 shares regularity properties with the usual distance function to a compact. Geometric stability results in this measure-theoretic setting : topology/geometry of the sublevel sets of d µ,m0, stable notion of persistence diagram for µ,. Algorithm : for nite point clouds d µ,m0 and (d µ,m0 ) can be easily and eciently computed pointwise in any dimension. F. Chazal, D. Cohen-Steiner, Q. Mérigot (Geometrica) Inference for Probability distributions 2009/07/08 32 / 32

Geometric Inference for Probability distributions

Geometric Inference for Probability distributions Geometric Inference for Probability distributions F. Chazal 1 D. Cohen-Steiner 2 Q. Mérigot 2 1 Geometrica, INRIA Saclay, 2 Geometrica, INRIA Sophia-Antipolis 2009 June 1 Motivation What is the (relevant)

More information

A sampling theory for compact sets

A sampling theory for compact sets ENS Lyon January 2010 A sampling theory for compact sets F. Chazal Geometrica Group INRIA Saclay To download these slides: http://geometrica.saclay.inria.fr/team/fred.chazal/teaching/distancefunctions.pdf

More information

Stability of boundary measures

Stability of boundary measures Stability of boundary measures F. Chazal D. Cohen-Steiner Q. Mérigot INRIA Saclay - Ile de France LIX, January 2008 Point cloud geometry Given a set of points sampled near an unknown shape, can we infer

More information

Geometric Inference on Kernel Density Estimates

Geometric Inference on Kernel Density Estimates Geometric Inference on Kernel Density Estimates Bei Wang 1 Jeff M. Phillips 2 Yan Zheng 2 1 Scientific Computing and Imaging Institute, University of Utah 2 School of Computing, University of Utah Feb

More information

Geometric inference for measures based on distance functions The DTM-signature for a geometric comparison of metric-measure spaces from samples

Geometric inference for measures based on distance functions The DTM-signature for a geometric comparison of metric-measure spaces from samples Distance to Geometric inference for measures based on distance functions The - for a geometric comparison of metric-measure spaces from samples the Ohio State University wan.252@osu.edu 1/31 Geometric

More information

Some Topics in Computational Topology. Yusu Wang. AMS Short Course 2014

Some Topics in Computational Topology. Yusu Wang. AMS Short Course 2014 Some Topics in Computational Topology Yusu Wang AMS Short Course 2014 Introduction Much recent developments in computational topology Both in theory and in their applications E.g, the theory of persistence

More information

Some Topics in Computational Topology

Some Topics in Computational Topology Some Topics in Computational Topology Yusu Wang Ohio State University AMS Short Course 2014 Introduction Much recent developments in computational topology Both in theory and in their applications E.g,

More information

Geometric Inference for Probability Measures

Geometric Inference for Probability Measures Geometric Inference for Probability Measures Frédéric CHAZAL David COHEN-STEINER Quentin MÉRIGOT June 21, 2010 Abstract Data often comes in the form of a point cloud sampled from an unknown compact subset

More information

Witnessed k-distance

Witnessed k-distance Witnessed k-distance Leonidas J. Guibas, Quentin Mérigot, Dmitriy Morozov To cite this version: Leonidas J. Guibas, Quentin Mérigot, Dmitriy Morozov. Witnessed k-distance. Discrete and Computational Geometry,

More information

WITNESSED K-DISTANCE

WITNESSED K-DISTANCE WITNESSED K-DISTANCE LEONIDAS GUIBAS, DMITRIY MOROZOV, AND QUENTIN MÉRIGOT Abstract. Distance function to a compact set plays a central role in several areas of computational geometry. Methods that rely

More information

Metric Thickenings of Euclidean Submanifolds

Metric Thickenings of Euclidean Submanifolds Metric Thickenings of Euclidean Submanifolds Advisor: Dr. Henry Adams Committe: Dr. Chris Peterson, Dr. Daniel Cooley Joshua Mirth Masters Thesis Defense October 3, 2017 Introduction Motivation Can we

More information

Persistent Homology of Data, Groups, Function Spaces, and Landscapes.

Persistent Homology of Data, Groups, Function Spaces, and Landscapes. Persistent Homology of Data, Groups, Function Spaces, and Landscapes. Shmuel Weinberger Department of Mathematics University of Chicago William Benter Lecture City University of Hong Kong Liu Bie Ju Centre

More information

Optimal Transport: A Crash Course

Optimal Transport: A Crash Course Optimal Transport: A Crash Course Soheil Kolouri and Gustavo K. Rohde HRL Laboratories, University of Virginia Introduction What is Optimal Transport? The optimal transport problem seeks the most efficient

More information

Learning from persistence diagrams

Learning from persistence diagrams Learning from persistence diagrams Ulrich Bauer TUM July 7, 2015 ACAT final meeting IST Austria Joint work with: Jan Reininghaus, Stefan Huber, Roland Kwitt 1 / 28 ? 2 / 28 ? 2 / 28 3 / 28 3 / 28 3 / 28

More information

Statistics and Topological Data Analysis

Statistics and Topological Data Analysis Geometry Understanding in Higher Dimensions CHAIRE D INFORMATIQUE ET SCIENCES NUMÉRIQUES Collège de France - June 2017 Statistics and Topological Data Analysis Bertrand MICHEL Ecole Centrale de Nantes

More information

Topology of the set of singularities of a solution of the Hamilton-Jacobi Equation

Topology of the set of singularities of a solution of the Hamilton-Jacobi Equation Topology of the set of singularities of a solution of the Hamilton-Jacobi Equation Albert Fathi IAS Princeton March 15, 2016 In this lecture, a singularity for a locally Lipschitz real valued function

More information

Declutter and Resample: Towards parameter free denoising

Declutter and Resample: Towards parameter free denoising Declutter and Resample: Towards parameter free denoising Mickaël Buchet, Tamal K. Dey, Jiayuan Wang, Yusu Wang arxiv:1511.05479v2 [cs.cg] 27 Mar 2017 Abstract In many data analysis applications the following

More information

Contractive metrics for scalar conservation laws

Contractive metrics for scalar conservation laws Contractive metrics for scalar conservation laws François Bolley 1, Yann Brenier 2, Grégoire Loeper 34 Abstract We consider nondecreasing entropy solutions to 1-d scalar conservation laws and show that

More information

Gromov-Hausdorff Approximation of Filament Structure Using Reeb-type Graph

Gromov-Hausdorff Approximation of Filament Structure Using Reeb-type Graph Gromov-Hausdorff Approximation of Filament Structure Using Reeb-type Graph Frédéric Chazal and Ruqi Huang and Jian Sun July 26, 2014 Abstract In many real-world applications data appear to be sampled around

More information

Homology through Interleaving

Homology through Interleaving Notes by Tamal K. Dey, OSU 91 Homology through Interleaving 32 Concept of interleaving A discrete set P R k is assumed to be a sample of a set X R k, if it lies near to it which we can quantify with the

More information

Inference of curvature using tubular neighborhoods

Inference of curvature using tubular neighborhoods Inference of curvature using tubular neighborhoods Frédéric Chazal, David Cohen-Steiner, André Lieutier Quentin Mérigot, Boris Thibert Introduction Geometric inference deals with the problem of recovering

More information

Integral Jensen inequality

Integral Jensen inequality Integral Jensen inequality Let us consider a convex set R d, and a convex function f : (, + ]. For any x,..., x n and λ,..., λ n with n λ i =, we have () f( n λ ix i ) n λ if(x i ). For a R d, let δ a

More information

Regularizing objective functionals in semi-supervised learning

Regularizing objective functionals in semi-supervised learning Regularizing objective functionals in semi-supervised learning Dejan Slepčev Carnegie Mellon University February 9, 2018 1 / 47 References S,Thorpe, Analysis of p-laplacian regularization in semi-supervised

More information

The optimal partial transport problem

The optimal partial transport problem The optimal partial transport problem Alessio Figalli Abstract Given two densities f and g, we consider the problem of transporting a fraction m [0, min{ f L 1, g L 1}] of the mass of f onto g minimizing

More information

Notions such as convergent sequence and Cauchy sequence make sense for any metric space. Convergent Sequences are Cauchy

Notions such as convergent sequence and Cauchy sequence make sense for any metric space. Convergent Sequences are Cauchy Banach Spaces These notes provide an introduction to Banach spaces, which are complete normed vector spaces. For the purposes of these notes, all vector spaces are assumed to be over the real numbers.

More information

LECTURE 10: THE ATIYAH-GUILLEMIN-STERNBERG CONVEXITY THEOREM

LECTURE 10: THE ATIYAH-GUILLEMIN-STERNBERG CONVEXITY THEOREM LECTURE 10: THE ATIYAH-GUILLEMIN-STERNBERG CONVEXITY THEOREM Contents 1. The Atiyah-Guillemin-Sternberg Convexity Theorem 1 2. Proof of the Atiyah-Guillemin-Sternberg Convexity theorem 3 3. Morse theory

More information

3 (Due ). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure?

3 (Due ). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure? MA 645-4A (Real Analysis), Dr. Chernov Homework assignment 1 (Due ). Show that the open disk x 2 + y 2 < 1 is a countable union of planar elementary sets. Show that the closed disk x 2 + y 2 1 is a countable

More information

7. The Stability Theorem

7. The Stability Theorem Vin de Silva Pomona College CBMS Lecture Series, Macalester College, June 2017 The Persistence Stability Theorem Stability Theorem (Cohen-Steiner, Edelsbrunner, Harer 2007) The following transformation

More information

Gromov-Hausdorff Approximation of Filament Structure Using Reeb-type Graph

Gromov-Hausdorff Approximation of Filament Structure Using Reeb-type Graph Noname manuscript No. (will be inserted by the editor) Gromov-Hausdorff Approximation of Filament Structure Using Reeb-type Graph Frédéric Chazal Ruqi Huang Jian Sun Received: date / Accepted: date Abstract

More information

1.1. MEASURES AND INTEGRALS

1.1. MEASURES AND INTEGRALS CHAPTER 1: MEASURE THEORY In this chapter we define the notion of measure µ on a space, construct integrals on this space, and establish their basic properties under limits. The measure µ(e) will be defined

More information

Minimal time mean field games

Minimal time mean field games based on joint works with Samer Dweik and Filippo Santambrogio PGMO Days 2017 Session Mean Field Games and applications EDF Lab Paris-Saclay November 14th, 2017 LMO, Université Paris-Sud Université Paris-Saclay

More information

LECTURE 15: COMPLETENESS AND CONVEXITY

LECTURE 15: COMPLETENESS AND CONVEXITY LECTURE 15: COMPLETENESS AND CONVEXITY 1. The Hopf-Rinow Theorem Recall that a Riemannian manifold (M, g) is called geodesically complete if the maximal defining interval of any geodesic is R. On the other

More information

Consistency of Modularity Clustering on Random Geometric Graphs

Consistency of Modularity Clustering on Random Geometric Graphs Consistency of Modularity Clustering on Random Geometric Graphs Erik Davis The University of Arizona May 10, 2016 Outline Introduction to Modularity Clustering Pointwise Convergence Convergence of Optimal

More information

Efficient and Optimal Modal-set Estimation using knn graphs

Efficient and Optimal Modal-set Estimation using knn graphs Efficient and Optimal Modal-set Estimation using knn graphs Samory Kpotufe ORFE, Princeton University Based on various results with Sanjoy Dasgupta, Kamalika Chaudhuri, Ulrike von Luxburg, Heinrich Jiang

More information

SEPARABILITY AND COMPLETENESS FOR THE WASSERSTEIN DISTANCE

SEPARABILITY AND COMPLETENESS FOR THE WASSERSTEIN DISTANCE SEPARABILITY AND COMPLETENESS FOR THE WASSERSTEIN DISTANCE FRANÇOIS BOLLEY Abstract. In this note we prove in an elementary way that the Wasserstein distances, which play a basic role in optimal transportation

More information

Manifold Reconstruction in Arbitrary Dimensions using Witness Complexes

Manifold Reconstruction in Arbitrary Dimensions using Witness Complexes Manifold Reconstruction in Arbitrary Dimensions using Witness Complexes Jean-Daniel Boissonnat INRIA, Geometrica 2004, route des lucioles 06902 Sophia-Antipolis boissonn@sophia.inria.fr Leonidas J. Guibas

More information

Pairwise Comparison Dynamics for Games with Continuous Strategy Space

Pairwise Comparison Dynamics for Games with Continuous Strategy Space Pairwise Comparison Dynamics for Games with Continuous Strategy Space Man-Wah Cheung https://sites.google.com/site/jennymwcheung University of Wisconsin Madison Department of Economics Nov 5, 2013 Evolutionary

More information

Large deviations for random projections of l p balls

Large deviations for random projections of l p balls 1/32 Large deviations for random projections of l p balls Nina Gantert CRM, september 5, 2016 Goal: Understanding random projections of high-dimensional convex sets. 2/32 2/32 Outline Goal: Understanding

More information

An introduction to Mathematical Theory of Control

An introduction to Mathematical Theory of Control An introduction to Mathematical Theory of Control Vasile Staicu University of Aveiro UNICA, May 2018 Vasile Staicu (University of Aveiro) An introduction to Mathematical Theory of Control UNICA, May 2018

More information

VARIATIONAL PRINCIPLE FOR THE ENTROPY

VARIATIONAL PRINCIPLE FOR THE ENTROPY VARIATIONAL PRINCIPLE FOR THE ENTROPY LUCIAN RADU. Metric entropy Let (X, B, µ a measure space and I a countable family of indices. Definition. We say that ξ = {C i : i I} B is a measurable partition if:

More information

Displacement convexity of the relative entropy in the discrete h

Displacement convexity of the relative entropy in the discrete h Displacement convexity of the relative entropy in the discrete hypercube LAMA Université Paris Est Marne-la-Vallée Phenomena in high dimensions in geometric analysis, random matrices, and computational

More information

Notes, March 4, 2013, R. Dudley Maximum likelihood estimation: actual or supposed

Notes, March 4, 2013, R. Dudley Maximum likelihood estimation: actual or supposed 18.466 Notes, March 4, 2013, R. Dudley Maximum likelihood estimation: actual or supposed 1. MLEs in exponential families Let f(x,θ) for x X and θ Θ be a likelihood function, that is, for present purposes,

More information

Vector spaces. DS-GA 1013 / MATH-GA 2824 Optimization-based Data Analysis.

Vector spaces. DS-GA 1013 / MATH-GA 2824 Optimization-based Data Analysis. Vector spaces DS-GA 1013 / MATH-GA 2824 Optimization-based Data Analysis http://www.cims.nyu.edu/~cfgranda/pages/obda_fall17/index.html Carlos Fernandez-Granda Vector space Consists of: A set V A scalar

More information

Techinical Proofs for Nonlinear Learning using Local Coordinate Coding

Techinical Proofs for Nonlinear Learning using Local Coordinate Coding Techinical Proofs for Nonlinear Learning using Local Coordinate Coding 1 Notations and Main Results Denition 1.1 (Lipschitz Smoothness) A function f(x) on R d is (α, β, p)-lipschitz smooth with respect

More information

THESIS METRIC THICKENINGS OF EUCLIDEAN SUBMANIFOLDS. Submitted by. Joshua R. Mirth. Department of Mathematics

THESIS METRIC THICKENINGS OF EUCLIDEAN SUBMANIFOLDS. Submitted by. Joshua R. Mirth. Department of Mathematics THESIS METRIC THICKENINGS OF EUCLIDEAN SUBMANIFOLDS Submitted by Joshua R. Mirth Department of Mathematics In partial fulfillment of the requirements For the Degree of Master of Science Colorado State

More information

Neighboring feasible trajectories in infinite dimension

Neighboring feasible trajectories in infinite dimension Neighboring feasible trajectories in infinite dimension Marco Mazzola Université Pierre et Marie Curie (Paris 6) H. Frankowska and E. M. Marchini Control of State Constrained Dynamical Systems Padova,

More information

Sobolev Mappings between Manifolds and Metric Spaces

Sobolev Mappings between Manifolds and Metric Spaces Sobolev Mappings between Manifolds and Metric Spaces Piotr Haj lasz Abstract In connection with the theory of p-harmonic mappings, Eells and Lemaire raised a question about density of smooth mappings in

More information

Lecture 1. Toric Varieties: Basics

Lecture 1. Toric Varieties: Basics Lecture 1. Toric Varieties: Basics Taras Panov Lomonosov Moscow State University Summer School Current Developments in Geometry Novosibirsk, 27 August1 September 2018 Taras Panov (Moscow University) Lecture

More information

Integration on Measure Spaces

Integration on Measure Spaces Chapter 3 Integration on Measure Spaces In this chapter we introduce the general notion of a measure on a space X, define the class of measurable functions, and define the integral, first on a class of

More information

Persistence based Clustering

Persistence based Clustering July 9, 2009 TGDA Workshop Persistence based Clustering Primoz Skraba joint work with Frédéric Chazal, Steve Y. Oudot, Leonidas J. Guibas 1-1 Clustering Input samples 2-1 Clustering Input samples Important

More information

Parameter-free Topology Inference and Sparsification for Data on Manifolds

Parameter-free Topology Inference and Sparsification for Data on Manifolds Parameter-free Topology Inference and Sparsification for Data on Manifolds Tamal K. Dey Zhe Dong Yusu Wang Abstract In topology inference from data, current approaches face two major problems. One concerns

More information

with Perfect Discrete Time Observations Campus de Beaulieu for all z 2 R d M z = h?1 (z) = fx 2 R m : h(x) = zg :

with Perfect Discrete Time Observations Campus de Beaulieu for all z 2 R d M z = h?1 (z) = fx 2 R m : h(x) = zg : Nonlinear Filtering with Perfect Discrete Time Observations Marc Joannides Department of Mathematics Imperial College London SW7 2B, United Kingdom mjo@ma.ic.ac.uk Francois LeGland IRISA / INRIA Campus

More information

Some Background Material

Some Background Material Chapter 1 Some Background Material In the first chapter, we present a quick review of elementary - but important - material as a way of dipping our toes in the water. This chapter also introduces important

More information

Optimization and Optimal Control in Banach Spaces

Optimization and Optimal Control in Banach Spaces Optimization and Optimal Control in Banach Spaces Bernhard Schmitzer October 19, 2017 1 Convex non-smooth optimization with proximal operators Remark 1.1 (Motivation). Convex optimization: easier to solve,

More information

PATH FUNCTIONALS OVER WASSERSTEIN SPACES. Giuseppe Buttazzo. Dipartimento di Matematica Università di Pisa.

PATH FUNCTIONALS OVER WASSERSTEIN SPACES. Giuseppe Buttazzo. Dipartimento di Matematica Università di Pisa. PATH FUNCTIONALS OVER WASSERSTEIN SPACES Giuseppe Buttazzo Dipartimento di Matematica Università di Pisa buttazzo@dm.unipi.it http://cvgmt.sns.it ENS Ker-Lann October 21-23, 2004 Several natural structures

More information

Interleaving Distance between Merge Trees

Interleaving Distance between Merge Trees Noname manuscript No. (will be inserted by the editor) Interleaving Distance between Merge Trees Dmitriy Morozov Kenes Beketayev Gunther H. Weber the date of receipt and acceptance should be inserted later

More information

18.175: Lecture 3 Integration

18.175: Lecture 3 Integration 18.175: Lecture 3 Scott Sheffield MIT Outline Outline Recall definitions Probability space is triple (Ω, F, P) where Ω is sample space, F is set of events (the σ-algebra) and P : F [0, 1] is the probability

More information

Mañé s Conjecture from the control viewpoint

Mañé s Conjecture from the control viewpoint Mañé s Conjecture from the control viewpoint Université de Nice - Sophia Antipolis Setting Let M be a smooth compact manifold of dimension n 2 be fixed. Let H : T M R be a Hamiltonian of class C k, with

More information

THEOREMS, ETC., FOR MATH 515

THEOREMS, ETC., FOR MATH 515 THEOREMS, ETC., FOR MATH 515 Proposition 1 (=comment on page 17). If A is an algebra, then any finite union or finite intersection of sets in A is also in A. Proposition 2 (=Proposition 1.1). For every

More information

2 (Bonus). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure?

2 (Bonus). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure? MA 645-4A (Real Analysis), Dr. Chernov Homework assignment 1 (Due 9/5). Prove that every countable set A is measurable and µ(a) = 0. 2 (Bonus). Let A consist of points (x, y) such that either x or y is

More information

Generative Models and Optimal Transport

Generative Models and Optimal Transport Generative Models and Optimal Transport Marco Cuturi Joint work / work in progress with G. Peyré, A. Genevay (ENS), F. Bach (INRIA), G. Montavon, K-R Müller (TU Berlin) Statistics 0.1 : Density Fitting

More information

g 2 (x) (1/3)M 1 = (1/3)(2/3)M.

g 2 (x) (1/3)M 1 = (1/3)(2/3)M. COMPACTNESS If C R n is closed and bounded, then by B-W it is sequentially compact: any sequence of points in C has a subsequence converging to a point in C Conversely, any sequentially compact C R n is

More information

Lecture Notes 1: Vector spaces

Lecture Notes 1: Vector spaces Optimization-based data analysis Fall 2017 Lecture Notes 1: Vector spaces In this chapter we review certain basic concepts of linear algebra, highlighting their application to signal processing. 1 Vector

More information

The uniformly accelerated motion in General Relativity from a geometric point of view. 1. Introduction. Daniel de la Fuente

The uniformly accelerated motion in General Relativity from a geometric point of view. 1. Introduction. Daniel de la Fuente XI Encuentro Andaluz de Geometría IMUS (Universidad de Sevilla), 15 de mayo de 2015, págs. 2934 The uniformly accelerated motion in General Relativity from a geometric point of view Daniel de la Fuente

More information

4 Countability axioms

4 Countability axioms 4 COUNTABILITY AXIOMS 4 Countability axioms Definition 4.1. Let X be a topological space X is said to be first countable if for any x X, there is a countable basis for the neighborhoods of x. X is said

More information

Master of Arts In Mathematics

Master of Arts In Mathematics ESTIMATING THE FRACTAL DIMENSION OF SETS DETERMINED BY NONERGODIC PARAMETERS A thesis submitted to the faculty of San Francisco State University In partial fulllment of The Requirements for The Degree

More information

(U) =, if 0 U, 1 U, (U) = X, if 0 U, and 1 U. (U) = E, if 0 U, but 1 U. (U) = X \ E if 0 U, but 1 U. n=1 A n, then A M.

(U) =, if 0 U, 1 U, (U) = X, if 0 U, and 1 U. (U) = E, if 0 U, but 1 U. (U) = X \ E if 0 U, but 1 U. n=1 A n, then A M. 1. Abstract Integration The main reference for this section is Rudin s Real and Complex Analysis. The purpose of developing an abstract theory of integration is to emphasize the difference between the

More information

Distribution-specific analysis of nearest neighbor search and classification

Distribution-specific analysis of nearest neighbor search and classification Distribution-specific analysis of nearest neighbor search and classification Sanjoy Dasgupta University of California, San Diego Nearest neighbor The primeval approach to information retrieval and classification.

More information

II - REAL ANALYSIS. This property gives us a way to extend the notion of content to finite unions of rectangles: we define

II - REAL ANALYSIS. This property gives us a way to extend the notion of content to finite unions of rectangles: we define 1 Measures 1.1 Jordan content in R N II - REAL ANALYSIS Let I be an interval in R. Then its 1-content is defined as c 1 (I) := b a if I is bounded with endpoints a, b. If I is unbounded, we define c 1

More information

Scalar Field Analysis over Point Cloud Data

Scalar Field Analysis over Point Cloud Data DOI 10.1007/s00454-011-9360-x Scalar Field Analysis over Point Cloud Data Frédéric Chazal Leonidas J. Guibas Steve Y. Oudot Primoz Skraba Received: 19 July 2010 / Revised: 13 April 2011 / Accepted: 18

More information

Entrance Exam, Real Analysis September 1, 2017 Solve exactly 6 out of the 8 problems

Entrance Exam, Real Analysis September 1, 2017 Solve exactly 6 out of the 8 problems September, 27 Solve exactly 6 out of the 8 problems. Prove by denition (in ɛ δ language) that f(x) = + x 2 is uniformly continuous in (, ). Is f(x) uniformly continuous in (, )? Prove your conclusion.

More information

Distributed learning in potential games over large-scale networks

Distributed learning in potential games over large-scale networks Distributed learning in potential games over large-scale networks Fabio Fagnani, DISMA, Politecnico di Torino joint work with Giacomo Como, Lund University Sandro Zampieri, DEI, University of Padova Networking

More information

Real Analysis II, Winter 2018

Real Analysis II, Winter 2018 Real Analysis II, Winter 2018 From the Finnish original Moderni reaalianalyysi 1 by Ilkka Holopainen adapted by Tuomas Hytönen January 18, 2018 1 Version dated September 14, 2011 Contents 1 General theory

More information

Quadratic estimates and perturbations of Dirac type operators on doubling measure metric spaces

Quadratic estimates and perturbations of Dirac type operators on doubling measure metric spaces Quadratic estimates and perturbations of Dirac type operators on doubling measure metric spaces Lashi Bandara maths.anu.edu.au/~bandara Mathematical Sciences Institute Australian National University February

More information

Coarse Geometry. Phanuel Mariano. Fall S.i.g.m.a. Seminar. Why Coarse Geometry? Coarse Invariants A Coarse Equivalence to R 1

Coarse Geometry. Phanuel Mariano. Fall S.i.g.m.a. Seminar. Why Coarse Geometry? Coarse Invariants A Coarse Equivalence to R 1 Coarse Geometry 1 1 University of Connecticut Fall 2014 - S.i.g.m.a. Seminar Outline 1 Motivation 2 3 The partition space P ([a, b]). Preliminaries Main Result 4 Outline Basic Problem 1 Motivation 2 3

More information

Local regression, intrinsic dimension, and nonparametric sparsity

Local regression, intrinsic dimension, and nonparametric sparsity Local regression, intrinsic dimension, and nonparametric sparsity Samory Kpotufe Toyota Technological Institute - Chicago and Max Planck Institute for Intelligent Systems I. Local regression and (local)

More information

n [ F (b j ) F (a j ) ], n j=1(a j, b j ] E (4.1)

n [ F (b j ) F (a j ) ], n j=1(a j, b j ] E (4.1) 1.4. CONSTRUCTION OF LEBESGUE-STIELTJES MEASURES In this section we shall put to use the Carathéodory-Hahn theory, in order to construct measures with certain desirable properties first on the real line

More information

CONSISTENCY OF CHEEGER AND RATIO GRAPH CUTS

CONSISTENCY OF CHEEGER AND RATIO GRAPH CUTS CONSISTENCY OF CHEEGER AN RATIO GRAPH CUTS NICOLÁS GARCÍA TRILLOS 1, EJAN SLEPČEV 1, JAMES VON BRECHT 2, THOMAS LAURENT 3, XAVIER BRESSON 4. ABSTRACT. This paper establishes the consistency of a family

More information

Centre for Mathematics and Its Applications The Australian National University Canberra, ACT 0200 Australia. 1. Introduction

Centre for Mathematics and Its Applications The Australian National University Canberra, ACT 0200 Australia. 1. Introduction ON LOCALLY CONVEX HYPERSURFACES WITH BOUNDARY Neil S. Trudinger Xu-Jia Wang Centre for Mathematics and Its Applications The Australian National University Canberra, ACT 0200 Australia Abstract. In this

More information

generalized modes in bayesian inverse problems

generalized modes in bayesian inverse problems generalized modes in bayesian inverse problems Christian Clason Tapio Helin Remo Kretschmann Petteri Piiroinen June 1, 2018 Abstract This work is concerned with non-parametric modes and MAP estimates for

More information

RESEARCH STATEMENT MICHAEL MUNN

RESEARCH STATEMENT MICHAEL MUNN RESEARCH STATEMENT MICHAEL MUNN Ricci curvature plays an important role in understanding the relationship between the geometry and topology of Riemannian manifolds. Perhaps the most notable results in

More information

5.3. Compactness results for Dirichlet spaces Compactness results for manifolds 39 References 39 sec:intro 1. Introduction We mean by a spectr

5.3. Compactness results for Dirichlet spaces Compactness results for manifolds 39 References 39 sec:intro 1. Introduction We mean by a spectr CONVERGENCE OF SPECTRAL STRUCTURES: A FUNCTIONAL ANALYTIC THEORY AND ITS APPLICATIONS TO SPECTRAL GEOMETRY KAZUHIRO KUWAE AND TAKASHI SHIOYA Abstract. The purpose of this paper is to present a functional

More information

Generic Aubry sets on surfaces

Generic Aubry sets on surfaces Université de Nice - Sophia Antipolis & Institut Universitaire de France Nanjing University (June 10th, 2013) Setting Let M be a smooth compact manifold of dimension n 2 be fixed. Let H : T M R be a Hamiltonian

More information

Cones of measures. Tatiana Toro. University of Washington. Quantitative and Computational Aspects of Metric Geometry

Cones of measures. Tatiana Toro. University of Washington. Quantitative and Computational Aspects of Metric Geometry Cones of measures Tatiana Toro University of Washington Quantitative and Computational Aspects of Metric Geometry Based on joint work with C. Kenig and D. Preiss Tatiana Toro (University of Washington)

More information

A Crash Course in Topological Groups

A Crash Course in Topological Groups A Crash Course in Topological Groups Iian B. Smythe Department of Mathematics Cornell University Olivetti Club November 8, 2011 Iian B. Smythe (Cornell) Topological Groups Nov. 8, 2011 1 / 28 Outline 1

More information

Statistical Measures of Uncertainty in Inverse Problems

Statistical Measures of Uncertainty in Inverse Problems Statistical Measures of Uncertainty in Inverse Problems Workshop on Uncertainty in Inverse Problems Institute for Mathematics and Its Applications Minneapolis, MN 19-26 April 2002 P.B. Stark Department

More information

Robustness for a Liouville type theorem in exterior domains

Robustness for a Liouville type theorem in exterior domains Robustness for a Liouville type theorem in exterior domains Juliette Bouhours 1 arxiv:1207.0329v3 [math.ap] 24 Oct 2014 1 UPMC Univ Paris 06, UMR 7598, Laboratoire Jacques-Louis Lions, F-75005, Paris,

More information

AN INTRODUCTION TO VARIFOLDS DANIEL WESER. These notes are from a talk given in the Junior Analysis seminar at UT Austin on April 27th, 2018.

AN INTRODUCTION TO VARIFOLDS DANIEL WESER. These notes are from a talk given in the Junior Analysis seminar at UT Austin on April 27th, 2018. AN INTRODUCTION TO VARIFOLDS DANIEL WESER These notes are from a talk given in the Junior Analysis seminar at UT Austin on April 27th, 2018. 1 Introduction Why varifolds? A varifold is a measure-theoretic

More information

Measure and integration

Measure and integration Chapter 5 Measure and integration In calculus you have learned how to calculate the size of different kinds of sets: the length of a curve, the area of a region or a surface, the volume or mass of a solid.

More information

Congurations of periodic orbits for equations with delayed positive feedback

Congurations of periodic orbits for equations with delayed positive feedback Congurations of periodic orbits for equations with delayed positive feedback Dedicated to Professor Tibor Krisztin on the occasion of his 60th birthday Gabriella Vas 1 MTA-SZTE Analysis and Stochastics

More information

A few words about the MTW tensor

A few words about the MTW tensor A few words about the Ma-Trudinger-Wang tensor Université Nice - Sophia Antipolis & Institut Universitaire de France Salah Baouendi Memorial Conference (Tunis, March 2014) The Ma-Trudinger-Wang tensor

More information

Towards a Mathematical Theory of Super-resolution

Towards a Mathematical Theory of Super-resolution Towards a Mathematical Theory of Super-resolution Carlos Fernandez-Granda www.stanford.edu/~cfgranda/ Information Theory Forum, Information Systems Laboratory, Stanford 10/18/2013 Acknowledgements This

More information

Nearest neighbor classification in metric spaces: universal consistency and rates of convergence

Nearest neighbor classification in metric spaces: universal consistency and rates of convergence Nearest neighbor classification in metric spaces: universal consistency and rates of convergence Sanjoy Dasgupta University of California, San Diego Nearest neighbor The primeval approach to classification.

More information

ON A MEASURE THEORETIC AREA FORMULA

ON A MEASURE THEORETIC AREA FORMULA ON A MEASURE THEORETIC AREA FORMULA VALENTINO MAGNANI Abstract. We review some classical differentiation theorems for measures, showing how they can be turned into an integral representation of a Borel

More information

Localizing solutions of the Einstein equations

Localizing solutions of the Einstein equations Localizing solutions of the Einstein equations Richard Schoen UC, Irvine and Stanford University - General Relativity: A Celebration of the 100th Anniversary, IHP - November 20, 2015 Plan of Lecture The

More information

Geometry and the Kato square root problem

Geometry and the Kato square root problem Geometry and the Kato square root problem Lashi Bandara Centre for Mathematics and its Applications Australian National University 29 July 2014 Geometric Analysis Seminar Beijing International Center for

More information

Introduction to the theory of currents. Tien-Cuong Dinh and Nessim Sibony

Introduction to the theory of currents. Tien-Cuong Dinh and Nessim Sibony Introduction to the theory of currents Tien-Cuong Dinh and Nessim Sibony September 21, 2005 2 3 This course is an introduction to the theory of currents. We give here the main notions with examples, exercises

More information

Clustering. CSL465/603 - Fall 2016 Narayanan C Krishnan

Clustering. CSL465/603 - Fall 2016 Narayanan C Krishnan Clustering CSL465/603 - Fall 2016 Narayanan C Krishnan ckn@iitrpr.ac.in Supervised vs Unsupervised Learning Supervised learning Given x ", y " "%& ', learn a function f: X Y Categorical output classification

More information

Part V. 17 Introduction: What are measures and why measurable sets. Lebesgue Integration Theory

Part V. 17 Introduction: What are measures and why measurable sets. Lebesgue Integration Theory Part V 7 Introduction: What are measures and why measurable sets Lebesgue Integration Theory Definition 7. (Preliminary). A measure on a set is a function :2 [ ] such that. () = 2. If { } = is a finite

More information

Domains in metric measure spaces with boundary of positive mean curvature, and the Dirichlet problem for functions of least gradient

Domains in metric measure spaces with boundary of positive mean curvature, and the Dirichlet problem for functions of least gradient Domains in metric measure spaces with boundary of positive mean curvature, and the Dirichlet problem for functions of least gradient Panu Lahti Lukáš Malý Nageswari Shanmugalingam Gareth Speight June 22,

More information