DOCSLIB.ORG

Gradient Systems

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text

Load more

Ralph Chill, Eva Fasangovˇ a´ Gradient Systems – 13th International Internet Seminar – February 10, 2010 c by R. Chill, E. Faˇsangov´a Contents 0 Introduction ................................................... 1 1 Gradient systems in euclidean space .............................. 5 1.1 Thespace Rd .............................................. 5 1.2 Fr´echetderivative .............................. ............ 7 1.3 Euclideangradient............................... ........... 8 1.4 Euclideangradientsystems ........................ .......... 9 1.5 Algebraicequationsand steepest descent ............ ........... 12 1.6 Exercises ....................................... .......... 13 2 Gradient systems in finite dimensional space ...................... 15 2.1 Definitionofgradient ............................. .......... 17 2.2 Definitionofgradientsystem....................... .......... 18 2.3 Local existence for ordinary differential equations . ............. 18 2.4 Local and global existence for gradientsystems .. ..... .......... 21 2.5 Newton’smethod.................................. ......... 23 2.6 Exercises ....................................... .......... 25 3 Gradients in infinite dimensional space: the Laplace operator on a bounded interval ....................... 27 3.1 Definitionofgradient ............................. .......... 28 3.2 Gradientsofquadraticforms ....................... .......... 30 3.3 Sobolevspacesinonespacedimension ................ ........ 31 3.4 The Dirichlet-Laplace operator on a bounded interval . .......... 33 3.5 The Dirichlet-Laplace operator with multiplicative coefficient ..... 35 3.6 Exercises ....................................... .......... 37 4 Gradients in infinite dimensional space: the Laplace operator and the p-Laplace operator .................. 39 4.1 Sobolevspacesinhigherdimensions ................. ......... 39 4.2 TheDirichlet-Laplaceoperator .................... ........... 42 v vi Contents 4.3 The Dirichlet-p-Laplaceoperator ............................. 44 4.4 Anauxiliaryresult............................... ........... 46 4.5 OriginsoftheLaplaceoperator..................... .......... 49 4.6 Exercises ....................................... .......... 52 5 Bochner-Lebesgue and Bochner-Sobolev spaces ................... 55 5.1 TheBochnerintegral.............................. .......... 56 5.2 Bochner-Lebesguespaces .......................... ......... 58 5.3 Bochner-Sobolevspaces in one space dimension . ..... .. ........ 59 5.4 Exercises ....................................... .......... 63 6 Gradient systems in infinite dimensional spaces: existence and uniqueness of solutions ......................................... 65 6.1 Gradient systems in infinite dimensional spaces .. ..... .......... 65 6.2 Global existence and uniqueness of solutions for gradient systems withconvexenergy................................... ...... 66 6.3 Exercises ....................................... .......... 75 7 Diffusion equations ............................................. 77 7.1 Heatconduction.................................. .......... 77 7.2 Thereaction-diffusionequation................... ............ 80 7.3 The Perona-Malikmodelin imageprocessing........... ........ 81 7.4 Global existenceand uniquenessof solutions ......... .......... 82 7.5 Exercises ....................................... .......... 84 8 Gradient systems in infinite dimensional spaces: existence and uniqueness of solutions II ....................................... 87 8.1 Global existence and uniqueness of solutions for gradient systems withellipticenergy ................................. ........ 88 8.2 Exercises ....................................... .......... 97 9 Regularity of solutions .......................................... 99 9.1 Interpolation................................... ............ 99 9.2 Timeregularity–Angenent’strick .................. ..........101 9.3 Exercises ....................................... ..........104 10 Existence of local solutions ......................................107 10.1 The local inverse theorem and existence of local solutions of abstractdifferentialequations ...................... ..........108 10.2 Gradients of quadratic forms and identification of the trace space.. 109 10.3 Asemilineardiffusionequation................... ............112 10.4 Exercises ...................................... ...........117 Contents vii 11 Asymptotic behaviour of solutions of gradient systems .............119 11.1 The ω-limit set of a continuous function on R+ .................120 11.2 A stabilisation result for global solutions of gradientsystems ......121 11.3 Nonstabilisation results for global solutions of gradient systems. 123 11.4 Exercises ...................................... ...........126 12 Asymptotic behaviour of solutions of gradient systems II ...........129 12.1 The Łojasiewicz-Simon inequality and stabilisation of global solutionsofgradientsystems......................... ........130 12.2 The Łojasiewicz-Simon inequality in Hilbert spaces . ...........133 12.3 Exercises ...................................... ...........138 13 The universe, soap bubbles and curve shortening ..................141 References ................................................... ......151 A Primerontopology .............................................1001 A.1 Metricspaces.................................... ..........1001 A.2 Sequences,convergence........................... ..........1004 A.3 Compactness ..................................... .........1006 A.4 Continuity ...................................... ..........1007 A.5 Completionofametricspace........................ .........1008 B Banach spaces and bounded linear operators ......................1011 B.1 Normedspaces.................................... .........1011 B.2 Productspacesandquotientspaces.................. ..........1017 B.3 Boundedlinearoperators .......................... ..........1020 B.4 TheArzela-Ascolitheorem ......................... .........1025 C Calculus on Banach spaces ......................................1029 C.1 Differentiable functions between Banach spaces . ...........1029 C.2 Local inverse theorem and implicit function theorem . ..........1030 C.3 *Newton’smethod................................. ........1034 D Hilbert spaces .................................................1035 D.1 Innerproductspaces .............................. ..........1035 D.2 Orthogonaldecomposition......................... ..........1039 D.3 *Fourierseries .................................. ..........1044 D.4 LinearfunctionalsonHilbertspaces................ ...........1048 D.5 WeakconvergenceinHilbertspaces .................. .........1049 E Dual spaces and weak convergence ...............................1053 E.1 ThetheoremofHahn-Banach ......................... .......1053 E.2 Weak∗ convergenceand the theorem of Banach-Alaoglu . .1059 E.3 Weakconvergenceandreflexivity .................... .........1060 E.4 *Minimizationofconvexfunctionals................ ..........1065 viii Contents E.5 *ThevonNeumannminimaxtheorem.................... .....1068 F The divergence theorem (BY WOLFGANG ARENDT) ................1071 F.1 Open sets of class C1 .......................................1071 F.2 Thedivergencetheorem ............................ .........1074 F.3 Proofofthedivergencetheorem ..................... .........1077 G Sobolev spaces .................................................1083 G.1 Test functions, convolution and regularization . .............1083 G.2 Sobolevspacesinonedimension ..................... ........1086 G.3 Sobolevspacesinseveraldimensions................ ..........1092 G.4 * Elliptic partial differentialequations ..... ..... ..............1094 H Bochner-Lebesgue and Bochner-Sobolev spaces ...................1097 H.1 TheBochnerintegral.............................. ..........1097 H.2 Bochner-Lebesguespaces .......................... .........1100 H.3 Bochner-Sobolevspaces in onedimension............. .........1102 Index ................................................... ..........1105 Lecture 0 Introduction Philosophy is written in that gigantic book which is perpetually open in front of our eyes (I allude to the universe), but no one can understand it who does not strive beforehand to learn the language and recognize the letters in which it is written. It is written in mathematical language, and its letters are triangles, circles and other geometrical figures, and without these means it is humanly impossible to understand any of it; without them, all we can do is to wander aimlessly in an obscure labyrinth. Galileo Galilei The evolution of many systems from physics, biology, chemistry, engineering, economics, social sciences can be rewritten in the mathematical language which uses ordinary or partial differential equations. Due to the analysis of the differen- tial equations we have the hope that we need not wander aimlessly in an obscure labyrinth but that we understand the corresponding evolution problems. In order to illustrate this statement, think of the motion of a pendulum, for example the oscillations of the great lantern in the nave of Pisa cathedral. At the end of the 16th century, after mere observation of this lantern and probably other pendula, Galilei conjectured that on the one hand the period of one swing is proportional to the length of the pendulum and on the other hand this period is independent of the amplitude of the swing. The conjecture about the precise period of one swing would have
Recommended publications
  • Arxiv:1507.07356V2 [Math.AP]

    Arxiv:1507.07356V2 [Math.AP]

    TEN EQUIVALENT DEFINITIONS OF THE FRACTIONAL LAPLACE OPERATOR MATEUSZ KWAŚNICKI Abstract. This article discusses several definitions of the fractional Laplace operator ( ∆)α/2 (α (0, 2)) in Rd (d 1), also known as the Riesz fractional derivative − ∈ ≥ operator, as an operator on Lebesgue spaces L p (p [1, )), on the space C0 of ∈ ∞ continuous functions vanishing at infinity and on the space Cbu of bounded uniformly continuous functions. Among these definitions are ones involving singular integrals, semigroups of operators, Bochner’s subordination and harmonic extensions. We collect and extend known results in order to prove that all these definitions agree: on each of the function spaces considered, the corresponding operators have common domain and they coincide on that common domain. 1. Introduction We consider the fractional Laplace operator L = ( ∆)α/2 in Rd, with α (0, 2) and d 1, 2, ... Numerous definitions of L can be found− in− literature: as a Fourier∈ multiplier with∈{ symbol} ξ α, as a fractional power in the sense of Bochner or Balakrishnan, as the inverse of the−| Riesz| potential operator, as a singular integral operator, as an operator associated to an appropriate Dirichlet form, as an infinitesimal generator of an appropriate semigroup of contractions, or as the Dirichlet-to-Neumann operator for an appropriate harmonic extension problem. Equivalence of these definitions for sufficiently smooth functions is well-known and easy. The purpose of this article is to prove that, whenever meaningful, all these definitions are equivalent in the Lebesgue space L p for p [1, ), ∈ ∞ in the space C0 of continuous functions vanishing at infinity, and in the space Cbu of bounded uniformly continuous functions.
  • The Matrix Calculus You Need for Deep Learning

    The Matrix Calculus You Need for Deep Learning

    The Matrix Calculus You Need For Deep Learning Terence Parr and Jeremy Howard July 3, 2018 (We teach in University of San Francisco's MS in Data Science program and have other nefarious projects underway. You might know Terence as the creator of the ANTLR parser generator. For more material, see Jeremy's fast.ai courses and University of San Francisco's Data Institute in- person version of the deep learning course.) HTML version (The PDF and HTML were generated from markup using bookish) Abstract This paper is an attempt to explain all the matrix calculus you need in order to understand the training of deep neural networks. We assume no math knowledge beyond what you learned in calculus 1, and provide links to help you refresh the necessary math where needed. Note that you do not need to understand this material before you start learning to train and use deep learning in practice; rather, this material is for those who are already familiar with the basics of neural networks, and wish to deepen their understanding of the underlying math. Don't worry if you get stuck at some point along the way|just go back and reread the previous section, and try writing down and working through some examples. And if you're still stuck, we're happy to answer your questions in the Theory category at forums.fast.ai. Note: There is a reference section at the end of the paper summarizing all the key matrix calculus rules and terminology discussed here. arXiv:1802.01528v3 [cs.LG] 2 Jul 2018 1 Contents 1 Introduction 3 2 Review: Scalar derivative rules4 3 Introduction to vector calculus and partial derivatives5 4 Matrix calculus 6 4.1 Generalization of the Jacobian .
  • A Brief Tour of Vector Calculus

    A Brief Tour of Vector Calculus

    A BRIEF TOUR OF VECTOR CALCULUS A. HAVENS Contents 0 Prelude ii 1 Directional Derivatives, the Gradient and the Del Operator 1 1.1 Conceptual Review: Directional Derivatives and the Gradient........... 1 1.2 The Gradient as a Vector Field............................ 5 1.3 The Gradient Flow and Critical Points ....................... 10 1.4 The Del Operator and the Gradient in Other Coordinates*............ 17 1.5 Problems........................................ 21 2 Vector Fields in Low Dimensions 26 2 3 2.1 General Vector Fields in Domains of R and R . 26 2.2 Flows and Integral Curves .............................. 31 2.3 Conservative Vector Fields and Potentials...................... 32 2.4 Vector Fields from Frames*.............................. 37 2.5 Divergence, Curl, Jacobians, and the Laplacian................... 41 2.6 Parametrized Surfaces and Coordinate Vector Fields*............... 48 2.7 Tangent Vectors, Normal Vectors, and Orientations*................ 52 2.8 Problems........................................ 58 3 Line Integrals 66 3.1 Defining Scalar Line Integrals............................. 66 3.2 Line Integrals in Vector Fields ............................ 75 3.3 Work in a Force Field................................. 78 3.4 The Fundamental Theorem of Line Integrals .................... 79 3.5 Motion in Conservative Force Fields Conserves Energy .............. 81 3.6 Path Independence and Corollaries of the Fundamental Theorem......... 82 3.7 Green's Theorem.................................... 84 3.8 Problems........................................ 89 4 Surface Integrals, Flux, and Fundamental Theorems 93 4.1 Surface Integrals of Scalar Fields........................... 93 4.2 Flux........................................... 96 4.3 The Gradient, Divergence, and Curl Operators Via Limits* . 103 4.4 The Stokes-Kelvin Theorem..............................108 4.5 The Divergence Theorem ...............................112 4.6 Problems........................................114 List of Figures 117 i 11/14/19 Multivariate Calculus: Vector Calculus Havens 0.
  • Curl, Divergence and Laplacian

    Curl, Divergence and Laplacian

    Curl, Divergence and Laplacian What to know: 1. The definition of curl and it two properties, that is, theorem 1, and be able to predict qualitatively how the curl of a vector field behaves from a picture. 2. The definition of divergence and it two properties, that is, if div F~ 6= 0 then F~ can't be written as the curl of another field, and be able to tell a vector field of clearly nonzero,positive or negative divergence from the picture. 3. Know the definition of the Laplace operator 4. Know what kind of objects those operator take as input and what they give as output. The curl operator Let's look at two plots of vector fields: Figure 1: The vector field Figure 2: The vector field h−y; x; 0i: h1; 1; 0i We can observe that the second one looks like it is rotating around the z axis. We'd like to be able to predict this kind of behavior without having to look at a picture. We also promised to find a criterion that checks whether a vector field is conservative in R3. Both of those goals are accomplished using a tool called the curl operator, even though neither of those two properties is exactly obvious from the definition we'll give. Definition 1. Let F~ = hP; Q; Ri be a vector field in R3, where P , Q and R are continuously differentiable. We define the curl operator: @R @Q @P @R @Q @P curl F~ = − ~i + − ~j + − ~k: (1) @y @z @z @x @x @y Remarks: 1.
  • Policy Gradient

    Policy Gradient

    Lecture 7: Policy Gradient Lecture 7: Policy Gradient David Silver Lecture 7: Policy Gradient Outline 1 Introduction 2 Finite Difference Policy Gradient 3 Monte-Carlo Policy Gradient 4 Actor-Critic Policy Gradient Lecture 7: Policy Gradient Introduction Policy-Based Reinforcement Learning In the last lecture we approximated the value or action-value function using parameters θ, V (s) V π(s) θ ≈ Q (s; a) Qπ(s; a) θ ≈ A policy was generated directly from the value function e.g. using -greedy In this lecture we will directly parametrise the policy πθ(s; a) = P [a s; θ] j We will focus again on model-free reinforcement learning Lecture 7: Policy Gradient Introduction Value-Based and Policy-Based RL Value Based Learnt Value Function Implicit policy Value Function Policy (e.g. -greedy) Policy Based Value-Based Actor Policy-Based No Value Function Critic Learnt Policy Actor-Critic Learnt Value Function Learnt Policy Lecture 7: Policy Gradient Introduction Advantages of Policy-Based RL Advantages: Better convergence properties Effective in high-dimensional or continuous action spaces Can learn stochastic policies Disadvantages: Typically converge to a local rather than global optimum Evaluating a policy is typically inefficient and high variance Lecture 7: Policy Gradient Introduction Rock-Paper-Scissors Example Example: Rock-Paper-Scissors Two-player game of rock-paper-scissors Scissors beats paper Rock beats scissors Paper beats rock Consider policies for iterated rock-paper-scissors A deterministic policy is easily exploited A uniform random policy
  • High Order Gradient, Curl and Divergence Conforming Spaces, with an Application to NURBS-Based Isogeometric Analysis

    High Order Gradient, Curl and Divergence Conforming Spaces, with an Application to NURBS-Based Isogeometric Analysis

    High order gradient, curl and divergence conforming spaces, with an application to compatible NURBS-based IsoGeometric Analysis R.R. Hiemstraa, R.H.M. Huijsmansa, M.I.Gerritsmab aDepartment of Marine Technology, Mekelweg 2, 2628CD Delft bDepartment of Aerospace Technology, Kluyverweg 2, 2629HT Delft Abstract Conservation laws, in for example, electromagnetism, solid and fluid mechanics, allow an exact discrete representation in terms of line, surface and volume integrals. We develop high order interpolants, from any basis that is a partition of unity, that satisfy these integral relations exactly, at cell level. The resulting gradient, curl and divergence conforming spaces have the propertythat the conservationlaws become completely independent of the basis functions. This means that the conservation laws are exactly satisfied even on curved meshes. As an example, we develop high ordergradient, curl and divergence conforming spaces from NURBS - non uniform rational B-splines - and thereby generalize the compatible spaces of B-splines developed in [1]. We give several examples of 2D Stokes flow calculations which result, amongst others, in a point wise divergence free velocity field. Keywords: Compatible numerical methods, Mixed methods, NURBS, IsoGeometric Analyis Be careful of the naive view that a physical law is a mathematical relation between previously defined quantities. The situation is, rather, that a certain mathematical structure represents a given physical structure. Burke [2] 1. Introduction In deriving mathematical models for physical theories, we frequently start with analysis on finite dimensional geometric objects, like a control volume and its bounding surfaces. We assign global, ’measurable’, quantities to these different geometric objects and set up balance statements.
  • Lower Bounds for the Ruelle Spectrum of Analytic Expanding Circle Maps

    Lower Bounds for the Ruelle Spectrum of Analytic Expanding Circle Maps

    LOWER BOUNDS FOR THE RUELLE SPECTRUM OF ANALYTIC EXPANDING CIRCLE MAPS OSCAR F. BANDTLOW AND FRED´ ERIC´ NAUD Abstract. We prove that there exists a dense set of analytic expanding maps of the circle for which the Ruelle eigenvalues enjoy exponential lower bounds. The proof combines potential theoretic techniques and explicit calculations for the spectrum of expanding Blaschke products. 1. Introduction and statement One of the basic problems of smooth ergodic theory is to investigate the asymptotic behaviour of mixing systems. In particular, there is interest in precise quantitative results on the rate of decay of correlations for smooth observables. In his seminal paper [17], Ruelle showed that the long time asymptotic behaviour of analytic hyperbolic systems can be understood in terms of the Ruelle spectrum, that is, the spectrum of certain operators, known as transfer operators in this context, acting on suitable Banach spaces of holomorphic functions. Surprisingly, even for the simplest systems like analytic expanding maps of the circle, very few quantitative results on the Ruelle spectrum are known so far. Let us be more precise. With T denoting the unit circle in the complex plane C, a map τ : T ! T is said to be analytic expanding if τ has a holomorphic extension to a neighbourhood of T and we have inf jτ 0(z)j > 1 : z2T The Ruelle spectrum is commonly defined as the spectrum of the Perron-Frobenius or transfer operator (on a suitable space of holomorphic functions) given locally by d X 0 (Lτ f)(z) := !(τ) φk(z)f(φk(z)) ; (1) k=1 where φk denotes the k-th local inverse branch of the covering map τ : T ! T, and ( +1 if τ is orientation preserving; arXiv:1605.06247v1 [math.DS] 20 May 2016 !(τ) = (2) −1 if τ is orientation reversing.
  • 1 Space Curves and Tangent Lines 2 Gradients and Tangent Planes

    1 Space Curves and Tangent Lines 2 Gradients and Tangent Planes

    CLASS NOTES for CHAPTER 4, Nonlinear Programming 1 Space Curves and Tangent Lines Recall that space curves are de¯ned by a set of parametric equations, x1(t) 2 x2(t) 3 r(t) = . 6 . 7 6 7 6 xn(t) 7 4 5 In Calc III, we might have written this a little di®erently, ~r(t) =< x(t); y(t); z(t) > but here we want to use n dimensions rather than two or three dimensions. The derivative and antiderivative of r with respect to t is done component- wise, x10 (t) x1(t) dt 2 x20 (t) 3 2 R x2(t) dt 3 r(t) = . ; R(t) = . 6 . 7 6 R . 7 6 7 6 7 6 xn0 (t) 7 6 xn(t) dt 7 4 5 4 5 And the local linear approximation to r(t) is alsoRdone componentwise. The tangent line (in n dimensions) can be written easily- the derivative at t = a is the direction of the curve, so the tangent line is given by: x1(a) x10 (a) 2 x2(a) 3 2 x20 (a) 3 L(t) = . + t . 6 . 7 6 . 7 6 7 6 7 6 xn(a) 7 6 xn0 (a) 7 4 5 4 5 In Class Exercise: Use Maple to plot the curve r(t) = [cos(t); sin(t); t]T ¼ and its tangent line at t = 2 . 2 Gradients and Tangent Planes Let f : Rn R. In this case, we can write: ! y = f(x1; x2; x3; : : : ; xn) 1 Note that any function that we wish to optimize must be of this form- It would not make sense to ¯nd the maximum of a function like a space curve; n dimensional coordinates are not well-ordered like the real line- so the fol- lowing statement would be meaningless: (3; 5) > (1; 2).
  • Integrating Gradients

    Integrating Gradients

    Integrating gradients 1 dimension The \gradient" of a function f(x) in one dimension (i.e., depending on only one variable) is just the derivative, f 0(x). We want to solve f 0(x) = k(x); (1) where k(x) is a known function. When we find a primitive function K(x) to k(x), the general form of f is K plus an arbitrary constant, f(x) = K(x) + C: (2) Example: With f 0(x) = 2=x we find f(x) = 2 ln x + C, where C is an undetermined constant. 2 dimensions We let the function depend on two variables, f(x; y). When the gradient rf is known, we have known functions k1 and k2 for the partial derivatives: @ f(x; y) = k (x; y); @x 1 @ f(x; y) = k (x; y): (3) @y 2 @f To solve this, we integrate one of them. To be specific, we here integrate @x over x. We find a primitive function with respect to x (thus keeping y constant) to k1 and call it k3. The general form of f will be k3 plus an arbitrary term whose x-derivative is zero. In other words, f(x; y) = k3(x; y) + B(y); (4) where B(y) is an unknown function of y. We have made some progress, because we have replaced an unknown function of two variables with another unknown function depending only on one variable. @f The best way to come further is not to integrate @y over y. That would give us a second unknown function, D(x).
  • Mean Value Theorem on Manifolds

    Mean Value Theorem on Manifolds

    MEAN VALUE THEOREMS ON MANIFOLDS Lei Ni Abstract We derive several mean value formulae on manifolds, generalizing the clas- sical one for harmonic functions on Euclidean spaces as well as the results of Schoen-Yau, Michael-Simon, etc, on curved Riemannian manifolds. For the heat equation a mean value theorem with respect to `heat spheres' is proved for heat equation with respect to evolving Riemannian metrics via a space-time consideration. Some new monotonicity formulae are derived. As applications of the new local monotonicity formulae, some local regularity theorems concerning Ricci flow are proved. 1. Introduction The mean value theorem for harmonic functions plays an central role in the theory of harmonic functions. In this article we discuss its generalization on manifolds and show how such generalizations lead to various monotonicity formulae. The main focuses of this article are the corresponding results for the parabolic equations, on which there have been many works, including [Fu, Wa, FG, GL1, E1], and the application of the new monotonicity formula to the study of Ricci flow. Let us start with the Watson's mean value formula [Wa] for the heat equation. Let U be a open subset of Rn (or a Riemannian manifold). Assume that u(x; t) is 2 a C solution to the heat equation in a parabolic region UT = U £ (0;T ). For any (x; t) de¯ne the `heat ball' by 8 9 jx¡yj2 < ¡ 4(t¡s) = e ¡n E(x; t; r) := (y; s) js · t; n ¸ r : : (4¼(t ¡ s)) 2 ; Then Z 1 jx ¡ yj2 u(x; t) = n u(y; s) 2 dy ds r E(x;t;r) 4(t ¡ s) for each E(x; t; r) ½ UT .
  • 4.1 Discrete Differential Geometry

    4.1 Discrete Differential Geometry

    Spring 2015 CSCI 599: Digital Geometry Processing 4.1 Discrete Differential Geometry Hao Li http://cs599.hao-li.com 1 Outline • Discrete Differential Operators • Discrete Curvatures • Mesh Quality Measures 2 Differential Operators on Polygons Differential Properties! • Surface is sufficiently differentiable • Curvatures → 2nd derivatives 3 Differential Operators on Polygons Differential Properties! • Surface is sufficiently differentiable • Curvatures → 2nd derivatives Polygonal Meshes! • Piecewise linear approximations of smooth surface • Focus on Discrete Laplace Beltrami Operator • Discrete differential properties defined over (x) N 4 Local Averaging Local Neighborhood ( x ) of a point !x N • often coincides with mesh vertex vi • n-ring neighborhood ( v ) or local geodesic ball Nn i 5 Local Averaging Local Neighborhood ( x ) of a point !x N • often coincides with mesh vertex vi • n-ring neighborhood ( v ) or local geodesic ball Nn i Neighborhood size! • Large: smoothing is introduced, stable to noise • Small: fine scale variation, sensitive to noise 6 Local Averaging: 1-Ring (x) N x Barycentric cell Voronoi cell Mixed Voronoi cell (barycenters/edgemidpoints) (circumcenters) (circumcenters/midpoint) tight error bound better approximation 7 Barycentric Cells Connect edge midpoints and triangle barycenters! • Simple to compute • Area is 1/3 o triangle areas • Slightly wrong for obtuse triangles 8 Mixed Cells Connect edge midpoints and! • Circumcenters for non-obtuse triangles • Midpoint of opposite edge for obtuse triangles • Better approximation,
  • A Doubly Nonlocal Laplace Operator and Its Connection to the Classical Laplacian Petronela Radu University of Nebraska - Lincoln, Pradu@Unl.Edu

    A Doubly Nonlocal Laplace Operator and Its Connection to the Classical Laplacian Petronela Radu University of Nebraska - Lincoln, [email protected]

    University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Faculty Publications, Department of Mathematics Mathematics, Department of 2019 A Doubly Nonlocal Laplace Operator and Its Connection to the Classical Laplacian Petronela Radu University of Nebraska - Lincoln, [email protected] Kelseys Wells University of Nebraska - Lincoln Follow this and additional works at: https://digitalcommons.unl.edu/mathfacpub Part of the Applied Mathematics Commons, and the Mathematics Commons Radu, Petronela and Wells, Kelseys, "A Doubly Nonlocal Laplace Operator and Its Connection to the Classical Laplacian" (2019). Faculty Publications, Department of Mathematics. 215. https://digitalcommons.unl.edu/mathfacpub/215 This Article is brought to you for free and open access by the Mathematics, Department of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Faculty Publications, Department of Mathematics by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. A DOUBLY NONLOCAL LAPLACE OPERATOR AND ITS CONNECTION TO THE CLASSICAL LAPLACIAN PETRONELA RADU, KELSEY WELLS Abstract. In this paper, motivated by the state-based peridynamic frame- work, we introduce a new nonlocal Laplacian that exhibits double nonlocality through the use of iterated integral operators. The operator introduces addi- tional degrees of flexibility that can allow for better representation of physical phenomena at different scales and in materials with different properties. We study mathematical properties of this state-based Laplacian, including connec- tions with other nonlocal and local counterparts. Finally, we obtain explicit rates of convergence for this doubly nonlocal operator to the classical Laplacian as the radii for the horizons of interaction kernels shrink to zero. 1. Introduction Physical phenomena that are beset by discontinuities in the solution, or in the domain, have been challenging to study in the context of classical partial differen- tial equations (PDEs).