Optimal Transport: A Crash Course
|
|
- Dora Dean
- 5 years ago
- Views:
Transcription
1 Optimal Transport: A Crash Course Soheil Kolouri and Gustavo K. Rohde HRL Laboratories, University of Virginia
2 Introduction
3 What is Optimal Transport? The optimal transport problem seeks the most efficient way of transforming one distribution of mass to another.
4 What is Optimal Transport? The problem was originally studied by Gaspard Monge in the 18 th century.
5 What is Optimal Transport? I Minimizing the transportation cost: Heavy mass shouldn t be transported far! S. Kolouri and G. K. Rohde OT Crash Course
6 Monge Formulation
7 Monge formulation and Transport Maps: A map, f : X Y, for measures µ and ν defined on spaces X and Y is called a transport map iff, f 1 (A) I 0 (x)dx = I 1 (y)dy (1) A
8 Monge formulation and Transport Maps: A map, f : X Y, for measures µ and ν defined on spaces X and Y is called a transport map iff, f 1 (A) I 0 (x)dx = I 1 (y)dy (1) A Find the optimal transport map f : X Y that minimizes the expected cost of transportation, M(µ, ν) = inf f MP c(x, f(x))i 0 (x)dx (2) X
9 Monge formulation and Transport Maps: In the majority of engineering applications the cost is the Euclidean distance, M(µ, ν) = inf f MP x f(x) 2 I 0 (x)dx X MP = {f : X Y I 0 (x)dx = I 1 (y)dy} (3) f 1 (A) A Note that the objective function and the constraint in Eq. (3) are both nonlinear with respect to f.
10 Monge formulation and Transport Maps: In the majority of engineering applications the cost is the Euclidean distance, M(µ, ν) = inf f MP x f(x) 2 I 0 (x)dx X MP = {f : X Y I 0 (x)dx = I 1 (y)dy} (3) f 1 (A) A Note that the objective function and the constraint in Eq. (3) are both nonlinear with respect to f. When f exists and it is differentiable, above constraint can be written in differential form as, MP = {f : X Y det(df(x))i 1 (f(x)) = I 0 (x), x X} (4)
11 A Transport Map May Not Exist: A transport map, f, exists only if µ is an absolutely continuous measure with compact support and c(x, f(x)) is convex. Here an example where the transport map does not exists:
12 A Transport Map May Not Exist: A transport map, f, exists only if µ is an absolutely continuous measure with compact support and c(x, f(x)) is convex. Here an example where the transport map does not exists: Monge formulation is not suitable for analyzing point cloud distributions or any particle like distributions. This is when Kantorovich s formulation comes to rescue!
13 Kantorovich Formulation
14 Kantorovich formulation and Transport Plans: Working on optimal allocation of scarce resources, Kantorovich revisited the optimal transport problem in 1942.
15 Kantorovich formulation and Transport Plans: Working on optimal allocation of scarce resources, Kantorovich revisited the optimal transport problem in 1942.
16 Kantorovich formulation and Transport Plans: A transport plan between measures µ and ν defined on X and Y is a probability measure γ X Y with marginals, γ(x, A) = ν(a), γ(b, Y ) = µ(b) (5)
17 Kantorovich formulation and Transport Plans: A transport plan between measures µ and ν defined on X and Y is a probability measure γ X Y with marginals, γ(x, A) = ν(a), γ(b, Y ) = µ(b) (5) Kantorovich then formulated the transport problem as finding the transport plan that minimizes the expected cost with respect to γ, K(µ, ν) = min γ Γ(µ,ν) c(x, y)dγ(x, y) X Y Γ(µ, ν) = {γ γ(a, Y ) = µ(a), γ(x, B) = ν(b)} (6)
18 Kantorovich formulation and Transport Plans: A transport plan between measures µ and ν defined on X and Y is a probability measure γ X Y with marginals, γ(x, A) = ν(a), γ(b, Y ) = µ(b) (5) Kantorovich then formulated the transport problem as finding the transport plan that minimizes the expected cost with respect to γ, K(µ, ν) = min γ Γ(µ,ν) c(x, y)dγ(x, y) X Y Γ(µ, ν) = {γ γ(a, Y ) = µ(a), γ(x, B) = ν(b)} (6) Note that unlike Monge s formulation, the objective function and the constraints in Kantorvich s formulation are linear with respect to γ.
19 Discrete Kantorovich formulation (Earth Mover s Distance) Let µ = N i=1 p iδ xi and µ = M j=1 q jδ yj, where δ xi is a Dirac measure, K(µ, ν) = min γ c(x i, y j )γ ij i j s.t. γ ij = p i, γ ij = q j, γ ij 0 (7) j i
20 Kantorovich vs. Monge The following relationship holds between Monge s and Kantorovich s formulation, K(µ, ν) M(µ, ν) (8) When an optimal transport map exists, f : X Y, the optimal transport plan and the optimal transport map are related through, X Y c(x, y)dγ(x, y) = c(x, f(x))dµ(x) (9) X
21 Existence and uniqueness Brenier s theorem Let c(x, y) = x y 2 and let µ be absolutely continuous with respect to the Lebesgue measure. Then, there exists a unique optimal transport map f : X Y such that, f 1 (A) which is characterized as, dµ(x) = dν(y) A f(x) = x ψ(x) = ( 1 2 x 2 ψ(x)) }{{} φ(x) (10) for some concave scalar function ψ. In other words, f is the gradient of a convex scalar function φ, and therefore it is curl free.
22 Transport-Based Metrics
23 p-wasserstein distance Let P p(ω) be the set of Borel probability measures with finite p th moment defined on a given metric space (Ω, d). The p-wasserstein metric, W p, for p 1 on P p(ω) is then defined as the optimal transport problem with the cost function c(x, y) = d p (x, y). Let µ and ν be in P p(ω), then, ( 1 W p(µ, ν) = min d p p (x, y)dγ(x, y)) (11) γ Γ(µ,ν) Ω Ω
24 p-wasserstein distance Let P p(ω) be the set of Borel probability measures with finite p th moment defined on a given metric space (Ω, d). The p-wasserstein metric, W p, for p 1 on P p(ω) is then defined as the optimal transport problem with the cost function c(x, y) = d p (x, y). Let µ and ν be in P p(ω), then, ( 1 W p(µ, ν) = min d p p (x, y)dγ(x, y)) (11) γ Γ(µ,ν) Ω Ω or equivalently when the optimal transport map, f, exists, ( 1 W p(µ, ν) = min d p p (x, f(x))dµ(x)). (12) f MP Ω
25 p-wasserstein distance Let P p(ω) be the set of Borel probability measures with finite p th moment defined on a given metric space (Ω, d). The p-wasserstein metric, W p, for p 1 on P p(ω) is then defined as the optimal transport problem with the cost function c(x, y) = d p (x, y). Let µ and ν be in P p(ω), then, ( 1 W p(µ, ν) = min d p p (x, y)dγ(x, y)) (11) γ Γ(µ,ν) Ω Ω or equivalently when the optimal transport map, f, exists, ( 1 W p(µ, ν) = min d p p (x, f(x))dµ(x)). (12) f MP Ω In most engineering applications Ω R d and d(x, y) = x y.
26 p-wasserstein for 1D probability measures For absolutely continuous one-dimensional probability measures µ and ν on R with positive probability density functions I 0 and I 1, the optimal transport map has a closed form solution.
27 p-wasserstein for 1D probability measures For absolutely continuous one-dimensional probability measures µ and ν on R with positive probability density functions I 0 and I 1, the optimal transport map has a closed form solution. Let F µ and F ν be the cumulative distribution functions, { Fµ(x) = µ((, x)) F ν(y) = ν((, y)) (13)
28 p-wasserstein for 1D probability measures For absolutely continuous one-dimensional probability measures µ and ν on R with positive probability density functions I 0 and I 1, the optimal transport map has a closed form solution. Let F µ and F ν be the cumulative distribution functions, { Fµ(x) = µ((, x)) F ν(y) = ν((, y)) (13) In one dimension, there only exists one monotonically increasing transport map f : R R such that f MP, and it is defined as, f(x) := min{t R : F ν(t) F µ(x)}. (14) or equivalently f(x) = Fν 1 F µ(x).
29 p-wasserstein distance for 1D probability measures 1 W p(µ, ν) = ( F 1 1 µ (t) Fν (t) p dt) p 1 (15) 0 Figure: Note that, the Euclidean distance does not provide a sensible distance between I 0, I 1 and I 2 while the p-wasserstein distance does.
30 p-wasserstein distance for 1D probability measures 1 W p(µ, ν) = ( F 1 1 µ (t) Fν (t) p dt) p 1 (16) 0 Figure: Note that, the Euclidean distance does not provide a sensible distance between I 0, I 1 and I 2 while the p-wasserstein distance does.
31 Sliced p-wasserstein distance Slice an n-dimensional probability distribution (n > 1) into one-dimensional representations through projections and measure p-wasserstein distance between these representations.
32 Sliced p-wasserstein distance Slice an n-dimensional probability distribution (n > 1) into one-dimensional representations through projections and measure p-wasserstein distance between these representations. Where R denotes Radon transform and is defined as, RI(t, θ) := S d 1 I(x)δ(t θ x)dx t R, θ S d 1 (Unit sphere in R d ) (17)
33 Sliced p-wasserstein distance Slice an n-dimensional probability distribution (n > 1) into one-dimensional representations through projections and measure p-wasserstein distance between these representations. Where R denotes Radon transform and is defined as, RI(t, θ) := S d 1 I(x)δ(t θ x)dx t R, θ S d 1 (Unit sphere in R d ) (17) and the p-sliced-wasserstein (p-sw) distance is defined as: ( ) 1 SW p(i 0, I 1 ) = W p S d 1 p (RI p 0(., θ), RI 1 (., θ))dθ (18)
34 Geometric Properties
35 2-Wasserstein geodesics The set of continuous measures together with the 2-Wasserstein metric forms a Riemmanian manifold. Given the 2-Wasserstein space, (P 2 (Ω), W 2 ), the geodesic between µ, ν P 2 (Ω) is the shortest curve on P 2 (Ω) that connects these measures.
36 2-Wasserstein geodesics The set of continuous measures together with the 2-Wasserstein metric forms a Riemmanian manifold. Given the 2-Wasserstein space, (P 2 (Ω), W 2 ), the geodesic between µ, ν P 2 (Ω) is the shortest curve on P 2 (Ω) that connects these measures. Let ρ t for t [0, 1] parametrizes a curve on P 2 (Ω) with ρ 0 = µ and ρ 1 = ν, and let I t denote the density of ρ t, I t(x)dx = dρ t(x).
37 2-Wasserstein geodesics The set of continuous measures together with the 2-Wasserstein metric forms a Riemmanian manifold. Given the 2-Wasserstein space, (P 2 (Ω), W 2 ), the geodesic between µ, ν P 2 (Ω) is the shortest curve on P 2 (Ω) that connects these measures. Let ρ t for t [0, 1] parametrizes a curve on P 2 (Ω) with ρ 0 = µ and ρ 1 = ν, and let I t denote the density of ρ t, I t(x)dx = dρ t(x). For the optimal transport map, f(x), between µ and ν the geodesic is parametrized as, I t(x) = det(df t(x))i 1 (f t(x)), f t(x) = (1 t)x + tf(x) (19)
38 2-Wasserstein geodesics The set of continuous measures together with the 2-Wasserstein metric forms a Riemmanian manifold. Given the 2-Wasserstein space, (P 2 (Ω), W 2 ), the geodesic between µ, ν P 2 (Ω) is the shortest curve on P 2 (Ω) that connects these measures. Let ρ t for t [0, 1] parametrizes a curve on P 2 (Ω) with ρ 0 = µ and ρ 1 = ν, and let I t denote the density of ρ t, I t(x)dx = dρ t(x). For the optimal transport map, f(x), between µ and ν the geodesic is parametrized as, I t(x) = det(df t(x))i 1 (f t(x)), f t(x) = (1 t)x + tf(x) (19) It is straightforward to show that, W 2 (µ, ρ t) = tw 2 (µ, ν) (20)
39 2-Wasserstein geodesics
40 2-Wasserstein geodesics
41 2-Wasserstein geodesics S. Kolouri and G. K. Rohde OT Crash Course
42 2-Wasserstein geodesics S. Kolouri and G. K. Rohde OT Crash Course
43 Numerical Solvers
44 Flow Minimization (Angenent, Haker, and Tannenbaum) The flow minimization method finds the optimal transport map following below steps: 1. Obtain an initial mass preserving transport map using the Knothe-Rosenblatt coupling 2. Update the initial map to obtain a curl free mass preserving transport map that minimizes the transport cost f k+1 = f k + ɛ 1 I 0 Df k (f k ( 1 div(f k ))), 1 : Poisson solver (21)
45 Flow Minimization (Angenent, Haker, and Tannenbaum) The flow minimization method finds the optimal transport map following below steps: 1. Obtain an initial mass preserving transport map using the Knothe-Rosenblatt coupling 2. Update the initial map to obtain a curl free mass preserving transport map that minimizes the transport cost f k+1 = f k + ɛ 1 I 0 Df k (f k ( 1 div(f k ))), 1 : Poisson solver (21) Angenent, S., et al. Minimizing flows for the Monge Kantorovich problem. SIAM 2003
46 Gradient descent on the dual problem (Chartrand et al.) I For the strictly convex cost function, c(x, y) = 1 x y 2, the dual of 2 Kantorovich problem can be formalized as minimizing, Z Z M (η) = η(x)dµ(x) + X η c (y)dν(y) Y η c (y) := maxx X (x y η(x)) is the Legendre-Fenchel transform of η(x). S. Kolouri and G. K. Rohde OT Crash Course (22)
47 Gradient descent on the dual problem (Chartrand et al.) I For the strictly convex cost function, c(x, y) = 1 x y 2, the dual of 2 Kantorovich problem can be formalized as minimizing, Z Z M (η) = η(x)dµ(x) + X η c (y)dν(y) (22) Y η c (y) := maxx X (x y η(x)) is the Legendre-Fenchel transform of η(x). I Then the potential transport field, η, is updated to minimize M (η) through, ηk+1 = ηk (I0 det(i Hη cc )I1 (id η cc )), S. Kolouri and G. K. Rohde OT Crash Course H: Hessian matrix (23)
48 Gradient descent on the dual problem (Chartrand et al.) I For the strictly convex cost function, c(x, y) = 1 x y 2, the dual of 2 Kantorovich problem can be formalized as minimizing, Z Z M (η) = η(x)dµ(x) + X η c (y)dν(y) (22) Y η c (y) := maxx X (x y η(x)) is the Legendre-Fenchel transform of η(x). I Then the potential transport field, η, is updated to minimize M (η) through, ηk+1 = ηk (I0 det(i Hη cc )I1 (id η cc )), H: Hessian matrix (23) Chartrand, R., et al. A gradient descent solution to the Monge-Kantorovich problem. AMS 2009 S. Kolouri and G. K. Rohde OT Crash Course
49 Linear programming Let µ = N i=1 p iδ xi and µ = M j=1 q jδ yj, where δ xi is a Dirac measure, K(µ, ν) = min γ c(x i, y j )γ ij i j s.t. γ ij = p i, γ ij = q j, γ ij 0 j i Can be solved through the Simplex algorithm or interior point techniques. Computational complexity of solvers: O(N 3 logn)
50 Linear programming Let µ = N i=1 p iδ xi and µ = M j=1 q jδ yj, where δ xi is a Dirac measure, K(µ, ν) = min γ c(x i, y j )γ ij i j s.t. γ ij = p i, γ ij = q j, γ ij 0 j i Can be solved through the Simplex algorithm or interior point techniques. Computational complexity of solvers: O(N 3 logn) Multi-Scale Approaches To improve computational complexity of several multi-scale approaches have been proposed The idea behind all these multi-scale techniques is to obtain a coarse transport plan and refine the transport plan iteratively.
51 Entropy Regularization I Cuturi proposed a regularized version of the Kantorovich problem which can be solved in O(N logn ), p (µ, ν) = Wp,λ Z dp (x, y)γ(x, y) + λγ(x, y) ln(γ(x, y))dxdy. inf γ Γ(µ,ν) Ω Ω S. Kolouri and G. K. Rohde OT Crash Course (24)
52 Entropy Regularization I Cuturi proposed a regularized version of the Kantorovich problem which can be solved in O(N logn ), p (µ, ν) = Wp,λ Z dp (x, y)γ(x, y) + λγ(x, y) ln(γ(x, y))dxdy. inf γ Γ(µ,ν) (24) Ω Ω I It is straightforward to show that the entropy regularized p-wasserstein distance in Equation (24) can be reformulated as, p Wp,λ (µ, ν) = λ inf γ Γ(µ,ν) KL(γ Kλ ), S. Kolouri and G. K. Rohde Kλ (x, y) = exp( OT Crash Course dp (x, y) ) λ (25)
53 Entropy Regularization I Cuturi proposed a regularized version of the Kantorovich problem which can be solved in O(N logn ), p (µ, ν) = Wp,λ Z dp (x, y)γ(x, y) + λγ(x, y) ln(γ(x, y))dxdy. inf γ Γ(µ,ν) (24) Ω Ω I It is straightforward to show that the entropy regularized p-wasserstein distance in Equation (24) can be reformulated as, p Wp,λ (µ, ν) = λ inf γ Γ(µ,ν) KL(γ Kλ ), Kλ (x, y) = exp( dp (x, y) ) λ (25) Cuturi, M. Sinkhorn distances: Lightspeed computation of optimal transport. NIPS S. Kolouri and G. K. Rohde OT Crash Course
54 Summary Introduction: 1. Monge formulation transport maps 2. Kantorovich formulation transport plans
55 Summary Introduction: 1. Monge formulation transport maps 2. Kantorovich formulation transport plans Transport-based metrics and geometric properties: 1. p-wasserstein metric 2. p-sliced Wasserstein metric 3. 2-Wasserstein geodesics
56 Summary Introduction: 1. Monge formulation transport maps 2. Kantorovich formulation transport plans Transport-based metrics and geometric properties: 1. p-wasserstein metric 2. p-sliced Wasserstein metric 3. 2-Wasserstein geodesics Numerical solvers: 1. Flow minimization (Monge) 2. Gradient descent on the dual problem (Monge) 3. Linear programming and multi-scale methods (Kantorovich) 4. Entropy regularized solver (Kantorovich)
57 Summary Introduction: 1. Monge formulation transport maps 2. Kantorovich formulation transport plans Transport-based metrics and geometric properties: 1. p-wasserstein metric 2. p-sliced Wasserstein metric 3. 2-Wasserstein geodesics Numerical solvers: 1. Flow minimization (Monge) 2. Gradient descent on the dual problem (Monge) 3. Linear programming and multi-scale methods (Kantorovich) 4. Entropy regularized solver (Kantorovich) Coming Up Next Transport-based transformations
58 Thank you!
Transport-based analysis, modeling, and learning from signal and data distributions
1 Transport-based analysis, modeling, and learning from signal and data distributions Soheil Kolouri, Serim Park, Matthew Thorpe, Dejan Slepčev, Gustavo K. Rohde arxiv:1609.04767v1 [cs.cv] 15 Sep 2016
More informationIntroduction to optimal transport
Introduction to optimal transport Nicola Gigli May 20, 2011 Content Formulation of the transport problem The notions of c-convexity and c-cyclical monotonicity The dual problem Optimal maps: Brenier s
More informationGenerative 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 informationOptimal Transport for Data Analysis
Optimal Transport for Data Analysis Bernhard Schmitzer 2017-05-16 1 Introduction 1.1 Reminders on Measure Theory Reference: Ambrosio, Fusco, Pallara: Functions of Bounded Variation and Free Discontinuity
More informationOptimal Transport for Applied Mathematicians
Optimal Transport for Applied Mathematicians Calculus of Variations, PDEs and Modelling Filippo Santambrogio 1 1 Laboratoire de Mathématiques d Orsay, Université Paris Sud, 91405 Orsay cedex, France filippo.santambrogio@math.u-psud.fr,
More informationOptimal Transport for Data Analysis
Optimal Transport for Data Analysis Bernhard Schmitzer 2017-05-30 1 Introduction 1.1 Reminders on Measure Theory Reference: Ambrosio, Fusco, Pallara: Functions of Bounded Variation and Free Discontinuity
More informationOptimal Transport and Wasserstein Distance
Optimal Transport and Wasserstein Distance The Wasserstein distance which arises from the idea of optimal transport is being used more and more in Statistics and Machine Learning. In these notes we review
More informationOptimal Transportation. Nonlinear Partial Differential Equations
Optimal Transportation and Nonlinear Partial Differential Equations Neil S. Trudinger Centre of Mathematics and its Applications Australian National University 26th Brazilian Mathematical Colloquium 2007
More informationThe 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 informationMoment Measures. Bo az Klartag. Tel Aviv University. Talk at the asymptotic geometric analysis seminar. Tel Aviv, May 2013
Tel Aviv University Talk at the asymptotic geometric analysis seminar Tel Aviv, May 2013 Joint work with Dario Cordero-Erausquin. A bijection We present a correspondence between convex functions and Borel
More informationA new Hellinger-Kantorovich distance between positive measures and optimal Entropy-Transport problems
A new Hellinger-Kantorovich distance between positive measures and optimal Entropy-Transport problems Giuseppe Savaré http://www.imati.cnr.it/ savare Dipartimento di Matematica, Università di Pavia Nonlocal
More informationNumerical Optimal Transport and Applications
Numerical Optimal Transport and Applications Gabriel Peyré Joint works with: Jean-David Benamou, Guillaume Carlier, Marco Cuturi, Luca Nenna, Justin Solomon www.numerical-tours.com Histograms in Imaging
More informationGradient Flows: Qualitative Properties & Numerical Schemes
Gradient Flows: Qualitative Properties & Numerical Schemes J. A. Carrillo Imperial College London RICAM, December 2014 Outline 1 Gradient Flows Models Gradient flows Evolving diffeomorphisms 2 Numerical
More informationRegularity for the optimal transportation problem with Euclidean distance squared cost on the embedded sphere
Regularity for the optimal transportation problem with Euclidean distance squared cost on the embedded sphere Jun Kitagawa and Micah Warren January 6, 011 Abstract We give a sufficient condition on initial
More informationsoftened probabilistic
Justin Solomon MIT Understand geometry from a softened probabilistic standpoint. Somewhere over here. Exactly here. One of these two places. Query 1 2 Which is closer, 1 or 2? Query 1 2 Which is closer,
More informationMATH MEASURE THEORY AND FOURIER ANALYSIS. Contents
MATH 3969 - MEASURE THEORY AND FOURIER ANALYSIS ANDREW TULLOCH Contents 1. Measure Theory 2 1.1. Properties of Measures 3 1.2. Constructing σ-algebras and measures 3 1.3. Properties of the Lebesgue measure
More informationIntroduction to Optimal Transport
Introduction to Optimal Transport Matthew Thorpe F2.08, Centre for Mathematical Sciences University of Cambridge Email: m.thorpe@maths.cam.ac.uk Lent 2018 Current Version: Thursday 8 th March, 2018 Foreword
More informationLogarithmic Sobolev Inequalities
Logarithmic Sobolev Inequalities M. Ledoux Institut de Mathématiques de Toulouse, France logarithmic Sobolev inequalities what they are, some history analytic, geometric, optimal transportation proofs
More informationGeometric 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 informationA description of transport cost for signed measures
A description of transport cost for signed measures Edoardo Mainini Abstract In this paper we develop the analysis of [AMS] about the extension of the optimal transport framework to the space of real measures.
More informationMeasure-theoretic probability
Measure-theoretic probability Koltay L. VEGTMAM144B November 28, 2012 (VEGTMAM144B) Measure-theoretic probability November 28, 2012 1 / 27 The probability space De nition The (Ω, A, P) measure space is
More informationA 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 informationOptimal mass transport as a distance measure between images
Optimal mass transport as a distance measure between images Axel Ringh 1 1 Department of Mathematics, KTH Royal Institute of Technology, Stockholm, Sweden. 21 st of June 2018 INRIA, Sophia-Antipolis, France
More informationProduct spaces and Fubini s theorem. Presentation for M721 Dec 7 th 2011 Chai Molina
Product spaces and Fubini s theorem Presentation for M721 Dec 7 th 2011 Chai Molina Motivation When can we exchange a double integral with iterated integrals? When are the two iterated integrals equal?
More informationDynamic and Stochastic Brenier Transport via Hopf-Lax formulae on Was
Dynamic and Stochastic Brenier Transport via Hopf-Lax formulae on Wasserstein Space With many discussions with Yann Brenier and Wilfrid Gangbo Brenierfest, IHP, January 9-13, 2017 ain points of the
More informationLocal semiconvexity of Kantorovich potentials on non-compact manifolds
Local semiconvexity of Kantorovich potentials on non-compact manifolds Alessio Figalli, Nicola Gigli Abstract We prove that any Kantorovich potential for the cost function c = d / on a Riemannian manifold
More informationDiscrete transport problems and the concavity of entropy
Discrete transport problems and the concavity of entropy Bristol Probability and Statistics Seminar, March 2014 Funded by EPSRC Information Geometry of Graphs EP/I009450/1 Paper arxiv:1303.3381 Motivating
More informationNishant Gurnani. GAN Reading Group. April 14th, / 107
Nishant Gurnani GAN Reading Group April 14th, 2017 1 / 107 Why are these Papers Important? 2 / 107 Why are these Papers Important? Recently a large number of GAN frameworks have been proposed - BGAN, LSGAN,
More informationGeometric 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 July 8 F. Chazal, D. Cohen-Steiner, Q.
More informationInformation geometry for bivariate distribution control
Information geometry for bivariate distribution control C.T.J.Dodson + Hong Wang Mathematics + Control Systems Centre, University of Manchester Institute of Science and Technology Optimal control of stochastic
More informationL. Decreusefond. Optimal transportation on configurations spaces
Optimal transportation on configurations spaces L. Decreusefond Monge problem µ and ν, two measures of same mass on R n. Find s : R n R n s.t. s µ = ν minimizing x s(x) dµ(x) Monge problem in R3 s x y
More informationConstrained Optimization and Lagrangian Duality
CIS 520: Machine Learning Oct 02, 2017 Constrained Optimization and Lagrangian Duality Lecturer: Shivani Agarwal Disclaimer: These notes are designed to be a supplement to the lecture. They may or may
More informationOptimal transportation on non-compact manifolds
Optimal transportation on non-compact manifolds Albert Fathi, Alessio Figalli 07 November 2007 Abstract In this work, we show how to obtain for non-compact manifolds the results that have already been
More informationGeodesic Density Tracking with Applications to Data Driven Modeling
Geodesic Density Tracking with Applications to Data Driven Modeling Abhishek Halder and Raktim Bhattacharya Department of Aerospace Engineering Texas A&M University College Station, TX 77843-3141 American
More informationStability 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 informationPoincaré Inequalities and Moment Maps
Tel-Aviv University Analysis Seminar at the Technion, Haifa, March 2012 Poincaré-type inequalities Poincaré-type inequalities (in this lecture): Bounds for the variance of a function in terms of the gradient.
More informationContinuum Modeling of Transportation Networks with Differential Equations
with Differential Equations King Abdullah University of Science and Technology Thuwal, KSA Examples of transportation networks The Silk Road Examples of transportation networks Painting by Latifa Echakhch
More informationFULL CHARACTERIZATION OF OPTIMAL TRANSPORT PLANS FOR CONCAVE COSTS
FULL CHARACTERIZATION OF OPTIMAL TRANSPORT PLANS FOR CONCAVE COSTS PAUL PEGON, DAVIDE PIAZZOLI, FILIPPO SANTAMBROGIO Abstract. This paper slightly improves a classical result by Gangbo and McCann (1996)
More information1 An Elementary Introduction to Monotone Transportation
An Elementary Introduction to Monotone Transportation after K. Ball [3] A summary written by Paata Ivanisvili Abstract We outline existence of the Brenier map. As an application we present simple proofs
More informationPATH 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 informationON THE EXISTENCE AND UNIQUENESS OF INVARIANT MEASURES ON LOCALLY COMPACT GROUPS. f(x) dx = f(y + a) dy.
N THE EXISTENCE AND UNIQUENESS F INVARIANT MEASURES N LCALLY CMPACT RUPS SIMN RUBINSTEIN-SALZED 1. Motivation and History ne of the most useful properties of R n is invariance under a linear transformation.
More informationOptimal transport for Seismic Imaging
Optimal transport for Seismic Imaging Bjorn Engquist In collaboration with Brittany Froese and Yunan Yang ICERM Workshop - Recent Advances in Seismic Modeling and Inversion: From Analysis to Applications,
More informationAPPLICATIONS DU TRANSPORT OPTIMAL DANS LES JEUX À POTENTIEL
APPLICATIONS DU TRANSPORT OPTIMAL DANS LES JEUX À POTENTIEL Adrien Blanchet IAST/TSE/Université Toulouse 1 Capitole Journées SMAI-MODE 216 En collaboration avec G. Carlier & P. Mossay & F. Santambrogio
More informationDisplacement 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 information9 Radon-Nikodym theorem and conditioning
Tel Aviv University, 2015 Functions of real variables 93 9 Radon-Nikodym theorem and conditioning 9a Borel-Kolmogorov paradox............. 93 9b Radon-Nikodym theorem.............. 94 9c Conditioning.....................
More informationarxiv: v2 [math.oc] 25 Sep 2012
From Knothe s Rearrangement to Brenier s Optimal Transport Map arxiv:1205.1099v2 [math.oc] 25 Sep 2012 Nicolas Bonnotte Université Paris-Sud, Orsay Scuola Normale Superiore, Pisa May 1, 2014 (revised version)
More informationOn the regularity of solutions of optimal transportation problems
On the regularity of solutions of optimal transportation problems Grégoire Loeper April 25, 2008 Abstract We give a necessary and sufficient condition on the cost function so that the map solution of Monge
More informationGradient Flow. Chang Liu. April 24, Tsinghua University. Chang Liu (THU) Gradient Flow April 24, / 91
Gradient Flow Chang Liu Tsinghua University April 24, 2017 Chang Liu (THU) Gradient Flow April 24, 2017 1 / 91 Contents 1 Introduction 2 Gradient flow in the Euclidean space Variants of Gradient Flow in
More informationFokker-Planck Equation on Graph with Finite Vertices
Fokker-Planck Equation on Graph with Finite Vertices January 13, 2011 Jointly with S-N Chow (Georgia Tech) Wen Huang (USTC) Hao-min Zhou(Georgia Tech) Functional Inequalities and Discrete Spaces Outline
More informationNotes of the seminar Evolution Equations in Probability Spaces and the Continuity Equation
Notes of the seminar Evolution Equations in Probability Spaces and the Continuity Equation Onno van Gaans Version of 12 April 2006 These are some notes supporting the seminar Evolution Equations in Probability
More informationTHESIS 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 informationConvex Optimization. Newton s method. ENSAE: Optimisation 1/44
Convex Optimization Newton s method ENSAE: Optimisation 1/44 Unconstrained minimization minimize f(x) f convex, twice continuously differentiable (hence dom f open) we assume optimal value p = inf x f(x)
More informationBayes spaces: use of improper priors and distances between densities
Bayes spaces: use of improper priors and distances between densities J. J. Egozcue 1, V. Pawlowsky-Glahn 2, R. Tolosana-Delgado 1, M. I. Ortego 1 and G. van den Boogaart 3 1 Universidad Politécnica de
More informationHessian Riemannian Gradient Flows in Convex Programming
Hessian Riemannian Gradient Flows in Convex Programming Felipe Alvarez, Jérôme Bolte, Olivier Brahic INTERNATIONAL CONFERENCE ON MODELING AND OPTIMIZATION MODOPT 2004 Universidad de La Frontera, Temuco,
More informationNumerical Approximation of L 1 Monge-Kantorovich Problems
Rome March 30, 2007 p. 1 Numerical Approximation of L 1 Monge-Kantorovich Problems Leonid Prigozhin Ben-Gurion University, Israel joint work with J.W. Barrett Imperial College, UK Rome March 30, 2007 p.
More informationFourier analysis, measures, and distributions. Alan Haynes
Fourier analysis, measures, and distributions Alan Haynes 1 Mathematics of diffraction Physical diffraction As a physical phenomenon, diffraction refers to interference of waves passing through some medium
More informationOptimal transport for Gaussian mixture models
1 Optimal transport for Gaussian mixture models Yongxin Chen, Tryphon T. Georgiou and Allen Tannenbaum Abstract arxiv:1710.07876v2 [math.pr] 30 Jan 2018 We present an optimal mass transport framework on
More informationminimize x subject to (x 2)(x 4) u,
Math 6366/6367: Optimization and Variational Methods Sample Preliminary Exam Questions 1. Suppose that f : [, L] R is a C 2 -function with f () on (, L) and that you have explicit formulae for
More informationOptimal Transport in ML
Optimal Transport in ML Rémi Gilleron Inria Lille & CRIStAL & Univ. Lille Feb. 2018 Main Source also some figures Computational Optimal transport, G. Peyré and M. Cuturi Rémi Gilleron (Magnet) OT Feb.
More informationarxiv: v3 [math.pr] 4 Sep 2018
Information Geometry manuscript No. (will be inserted by the editor) Logarithmic divergences from optimal transport and Rényi geometry Ting-Kam Leonard Wong arxiv:1712.03610v3 [math.pr] 4 Sep 2018 Received:
More informationOptimal Transport And WGAN
Optimal Transport And WGAN Yiping Lu 1 1 School of Mathmetical Science Peking University Data Science Seminar, PKU, 2017 Yiping Lu (pku) Geometry Of WGAN 1/34 Outline 1 Introduction To Optimal Transport.
More informationMTH 404: Measure and Integration
MTH 404: Measure and Integration Semester 2, 2012-2013 Dr. Prahlad Vaidyanathan Contents I. Introduction....................................... 3 1. Motivation................................... 3 2. The
More informationMurat Akman. The Brunn-Minkowski Inequality and a Minkowski Problem for Nonlinear Capacities. March 10
The Brunn-Minkowski Inequality and a Minkowski Problem for Nonlinear Capacities Murat Akman March 10 Postdoc in HA Group and Postdoc at the University of Connecticut Minkowski Addition of Sets Let E 1
More informationWasserstein Training of Boltzmann Machines
Wasserstein Training of Boltzmann Machines Grégoire Montavon, Klaus-Rober Muller, Marco Cuturi Presenter: Shiyu Liang December 1, 2016 Coordinated Science Laboratory Department of Electrical and Computer
More informationDeforming conformal metrics with negative Bakry-Émery Ricci Tensor on manifolds with boundary
Deforming conformal metrics with negative Bakry-Émery Ricci Tensor on manifolds with boundary Weimin Sheng (Joint with Li-Xia Yuan) Zhejiang University IMS, NUS, 8-12 Dec 2014 1 / 50 Outline 1 Prescribing
More informationSolutions to Tutorial 11 (Week 12)
THE UIVERSITY OF SYDEY SCHOOL OF MATHEMATICS AD STATISTICS Solutions to Tutorial 11 (Week 12) MATH3969: Measure Theory and Fourier Analysis (Advanced) Semester 2, 2017 Web Page: http://sydney.edu.au/science/maths/u/ug/sm/math3969/
More informationHyperbolic inverse problems and exact controllability
Hyperbolic inverse problems and exact controllability Lauri Oksanen University College London An inverse initial source problem Let M R n be a compact domain with smooth strictly convex boundary, and let
More informationSTEEPEST DESCENT FLOWS AND APPLICATIONS TO SPACES OF PROBABILITY MEASURES
1 STEEPEST DESCENT FLOWS AND APPLICATIONS TO SPACES OF PROBABILITY MEASURES LUIGI AMBROSIO LECTURE NOTES, SANTANDER (SPAIN), JULY 2004 1. Introduction These notes are the outcome of a series of seminars
More informationOptimal transport for machine learning
Optimal transport for machine learning Rémi Flamary AG GDR ISIS, Sète, 16 Novembre 2017 2 / 37 Collaborators N. Courty A. Rakotomamonjy D. Tuia A. Habrard M. Cuturi M. Perrot C. Févotte V. Emiya V. Seguy
More informationSMOOTH OPTIMAL TRANSPORTATION ON HYPERBOLIC SPACE
SMOOTH OPTIMAL TRANSPORTATION ON HYPERBOLIC SPACE JIAYONG LI A thesis completed in partial fulfilment of the requirement of Master of Science in Mathematics at University of Toronto. Copyright 2009 by
More informationSOLUTION OF POISSON S EQUATION. Contents
SOLUTION OF POISSON S EQUATION CRISTIAN E. GUTIÉRREZ OCTOBER 5, 2013 Contents 1. Differentiation under the integral sign 1 2. The Newtonian potential is C 1 2 3. The Newtonian potential from the 3rd Green
More informationPreconditioning techniques in frame theory and probabilistic frames
Preconditioning techniques in frame theory and probabilistic frames Department of Mathematics & Norbert Wiener Center University of Maryland, College Park AMS Short Course on Finite Frame Theory: A Complete
More informationarxiv: v1 [math.ap] 25 Oct 2012
DIMENSIONALITY OF LOCAL MINIMIZERS OF THE INTERACTION ENERGY D. BALAGUÉ 1, J. A. CARRILLO 2, T. LAURENT 3, AND G. RAOUL 4 arxiv:1210.6795v1 [math.ap] 25 Oct 2012 Abstract. In this work we consider local
More informationIntroduction to Optimal Transport Theory
Introduction to Optimal Transport Theory Filippo Santambrogio Grenoble, June 15th 2009 These very short lecture notes do not want to be an exhaustive presentation of the topic, but only a short list of
More informationGeneralized Orlicz spaces and Wasserstein distances for convex concave scale functions
Bull. Sci. math. 135 (2011 795 802 www.elsevier.com/locate/bulsci Generalized Orlicz spaces and Wasserstein distances for convex concave scale functions Karl-Theodor Sturm Institut für Angewandte Mathematik,
More informationJune 21, Peking University. Dual Connections. Zhengchao Wan. Overview. Duality of connections. Divergence: general contrast functions
Dual Peking University June 21, 2016 Divergences: Riemannian connection Let M be a manifold on which there is given a Riemannian metric g =,. A connection satisfying Z X, Y = Z X, Y + X, Z Y (1) for all
More informationBregman Divergences for Data Mining Meta-Algorithms
p.1/?? Bregman Divergences for Data Mining Meta-Algorithms Joydeep Ghosh University of Texas at Austin ghosh@ece.utexas.edu Reflects joint work with Arindam Banerjee, Srujana Merugu, Inderjit Dhillon,
More informationAdvanced Analysis Qualifying Examination Department of Mathematics and Statistics University of Massachusetts. Tuesday, January 16th, 2018
NAME: Advanced Analysis Qualifying Examination Department of Mathematics and Statistics University of Massachusetts Tuesday, January 16th, 2018 Instructions 1. This exam consists of eight (8) problems
More informationApproximations of displacement interpolations by entropic interpolations
Approximations of displacement interpolations by entropic interpolations Christian Léonard Université Paris Ouest Mokaplan 10 décembre 2015 Interpolations in P(X ) X : Riemannian manifold (state space)
More informationThe Monge problem on non-compact manifolds
The Monge problem on non-compact manifolds Alessio Figalli Scuola Normale Superiore Pisa Abstract In this paper we prove the existence of an optimal transport map on non-compact manifolds for a large class
More informationFrom Knothe s transport to Brenier s map and a continuation method for optimal transport
From Knothe s transport to Brenier s map and a continuation method for optimal transport G. Carlier, A. Galichon, F. Santambrogio November 4, 29 Abstract A simple procedure to map two probability measures
More informationThe semi-geostrophic equations - a model for large-scale atmospheric flows
The semi-geostrophic equations - a model for large-scale atmospheric flows Beatrice Pelloni, University of Reading with M. Cullen (Met Office), D. Gilbert, T. Kuna INI - MFE Dec 2013 Introduction - Motivation
More informationarxiv: v1 [math.ap] 3 Sep 2014
L 2 A NUMERICAL ALGORITHM FOR SEMI-DISCRETE OPTIMAL TRANSPORT IN 3D BRUNO LÉVY arxiv:1409.1279v1 [math.ap] 3 Sep 2014 Abstract. This paper introduces a numerical algorithm to compute the L 2 optimal transport
More informationMath 4121 Spring 2012 Weaver. Measure Theory. 1. σ-algebras
Math 4121 Spring 2012 Weaver Measure Theory 1. σ-algebras A measure is a function which gauges the size of subsets of a given set. In general we do not ask that a measure evaluate the size of every subset,
More informationREGULARITY OF POTENTIAL FUNCTIONS IN OPTIMAL TRANSPORTATION. Centre for Mathematics and Its Applications The Australian National University
ON STRICT CONVEXITY AND C 1 REGULARITY OF POTENTIAL FUNCTIONS IN OPTIMAL TRANSPORTATION Neil Trudinger Xu-Jia Wang Centre for Mathematics and Its Applications The Australian National University Abstract.
More informationOn the interior of the simplex, we have the Hessian of d(x), Hd(x) is diagonal with ith. µd(w) + w T c. minimize. subject to w T 1 = 1,
Math 30 Winter 05 Solution to Homework 3. Recognizing the convexity of g(x) := x log x, from Jensen s inequality we get d(x) n x + + x n n log x + + x n n where the equality is attained only at x = (/n,...,
More informationPenalized Barycenters in the Wasserstein space
Penalized Barycenters in the Wasserstein space Elsa Cazelles, joint work with Jérémie Bigot & Nicolas Papadakis Université de Bordeaux & CNRS Journées IOP - Du 5 au 8 Juillet 2017 Bordeaux Elsa Cazelles
More informationHÖLDER CONTINUITY FOR OPTIMAL MULTIVALUED MAPPINGS
HÖLDER CONTINUITY FOR OPTIMAL MULTIVALUED MAPPINGS R. J. MCCANN AND M. SOSIO Abstract. Gangbo and McCann showed that optimal transportation between hypersurfaces generally leads to multivalued optimal
More informationMathematical Foundations of Data Sciences
Mathematical Foundations of Data Sciences Gabriel Peyré CNRS & DMA École Normale Supérieure gabriel.peyre@ens.fr https://mathematical-tours.github.io www.numerical-tours.com January 7, 2018 Chapter 18
More information3. Convex functions. basic properties and examples. operations that preserve convexity. the conjugate function. quasiconvex functions
3. Convex functions Convex Optimization Boyd & Vandenberghe basic properties and examples operations that preserve convexity the conjugate function quasiconvex functions log-concave and log-convex functions
More informationConvex discretization of functionals involving the Monge-Ampère operator
Convex discretization of functionals involving the Monge-Ampère operator Quentin Mérigot CNRS / Université Paris-Dauphine Joint work with J.D. Benamou, G. Carlier and É. Oudet GeMeCod Conference October
More informationCurvature and the continuity of optimal transportation maps
Curvature and the continuity of optimal transportation maps Young-Heon Kim and Robert J. McCann Department of Mathematics, University of Toronto June 23, 2007 Monge-Kantorovitch Problem Mass Transportation
More informationON THE FOLIATION OF SPACE-TIME BY CONSTANT MEAN CURVATURE HYPERSURFACES
ON THE FOLIATION OF SPACE-TIME BY CONSTANT MEAN CURVATURE HYPERSURFACES CLAUS GERHARDT Abstract. We prove that the mean curvature τ of the slices given by a constant mean curvature foliation can be used
More informationOn Lagrangian solutions for the semi-geostrophic system with singular initial data
On Lagrangian solutions for the semi-geostrophic system with singular initial data Mikhail Feldman Department of Mathematics University of Wisconsin-Madison Madison, WI 5376, USA feldman@math.wisc.edu
More informationTransport Continuity Property
On Riemannian manifolds satisfying the Transport Continuity Property Université de Nice - Sophia Antipolis (Joint work with A. Figalli and C. Villani) I. Statement of the problem Optimal transport on Riemannian
More informationSome topics in sub-riemannian geometry
Some topics in sub-riemannian geometry Luca Rizzi CNRS, Institut Fourier Mathematical Colloquium Universität Bern - December 19 2016 Sub-Riemannian geometry Known under many names: Carnot-Carathéodory
More informationApplications of Linear Programming
Applications of Linear Programming lecturer: András London University of Szeged Institute of Informatics Department of Computational Optimization Lecture 9 Non-linear programming In case of LP, the goal
More informationA VARIATIONAL METHOD FOR THE ANALYSIS OF A MONOTONE SCHEME FOR THE MONGE-AMPÈRE EQUATION 1. INTRODUCTION
A VARIATIONAL METHOD FOR THE ANALYSIS OF A MONOTONE SCHEME FOR THE MONGE-AMPÈRE EQUATION GERARD AWANOU AND LEOPOLD MATAMBA MESSI ABSTRACT. We give a proof of existence of a solution to the discrete problem
More informationSeismic imaging and optimal transport
Seismic imaging and optimal transport Bjorn Engquist In collaboration with Brittany Froese, Sergey Fomel and Yunan Yang Brenier60, Calculus of Variations and Optimal Transportation, Paris, January 10-13,
More informationConvex discretization of functionals involving the Monge-Ampère operator
Convex discretization of functionals involving the Monge-Ampère operator Quentin Mérigot CNRS / Université Paris-Dauphine Joint work with J.D. Benamou, G. Carlier and É. Oudet Workshop on Optimal Transport
More information