Gradient flows of energy functionals on the space of probability measures with Wasserstein metric has proved to be a strong tool in studying certain mass conserving evolution equations. Such gradient flows provide an alternate formulation for the solutions of the corresponding evolution equations. An important condition, which is known to guarantees existence, uniqueness, and continuous dependence on initial data is that the corresponding energy functional be displacement convex. We introduce a relaxed notion of displacement convexity and we show that it still guarantees short time existence and uniqueness of Wasserstein gradient flows for higher order energy functionals which are not displacement convex in the standard sense. This extends the applicability of the gradient flow approach to larger family of energies. As an application, local and global well-posedness of different higher order non-linear evolution equations are derived. Examples include the thin-film equation and the quantum drift diffusion equation in one spatial variable.

Gradient flows of energy functionals on the space of probability measures with Wasserstein metric has proved to be a strong tool in studying certain mass conserving evolution equations. Such gradient flows provide an alternate formulation for the solutions of the corresponding evolution equations. An important condition, which is known to guarantees existence, uniqueness, and continuous dependence on initial data is that the corresponding energy functional be displacement convex. We introduce a relaxed notion of displacement convexity and we show that it still guarantees short time existence and uniqueness of Wasserstein gradient flows for higher order energy functionals which are not displacement convex in the standard sense. This extends the applicability of the gradient flow approach to larger family of energies. As an application, local and global well-posedness of different higher order non-linear evolution equations are derived. Examples include the thin-film equation and the quantum drift diffusion equation in one spatial variable.

This thesis makes significant contributions to the object detection problem in computer vision. The object detection problem is, given a digital image of a scene, to detect the relevant object in the image. One technique for performing object detection, called ``active contours,' optimizes a constructed energy that is defined on contours (closed curves) and is tailored to image features. An optimization method can be used to perform the optimization of the energy, and thereby deform an initially placed contour to the relevant object. The typical optimization technique used in almost every active contour paper is evolving the contour by the energy's gradient descent flow, i.e., the steepest descent flow, in order to drive the initial contour to (hopefully) the minimum curve. The problem with this technique is that often times the contour becomes stuck in a sub-optimal and undesirable local minimum of the energy. This problem can be partially attributed to the fact that the gradient flows of these energies make use of only local image and contour information. By local, we mean that in order to evolve a point on the contour, only information local to that point is used. Therefore, in this thesis, we introduce a new class of flows that are global in that the evolution of a point on the contour depends on global information from the entire curve. These flows help avoid a number of problems with traditional flows including helping in avoiding undesirable local minima. We demonstrate practical applications of these flows for the object detection problem, including applications to both image segmentation and visual object tracking.

In this thesis we study Sobolev gradient flows for Perona-Malik style energy functionals and generalizations thereof. We begin with first order isotropic flows which are shown to be regularizations of the heat equation. We show that these flows are well-posed in the forward and reverse directions which yields an effective linear sharpening algorithm. We furthermore establish a number of maximum principles for the forward flow and show that edges are preserved for a finite period of time. We then go on to study isotropic Sobolev gradient flows with respect to higher order Sobolev metrics. As the Sobolev order is increased, we observe an increasing reluctance to destroy fine details and texture. We then consider Sobolev gradient flows for non-linear anisotropic diffusion functionals of arbitrary order. We establish existence, uniqueness and continuous dependence on initial data for a broad class of such equations. The well-posedness of these new anisotropic gradient flows opens the door to a wide variety of sharpening and diffusion techniques which were previously impossible under L2 gradient descent. We show how one can easily use this framework to design an anisotropic sharpening algorithm which can sharpen image features while suppressing noise. We compare our sharpening algorithm to the well-known shock filter and show that Sobolev sharpening produces natural looking images without the "staircasing" artifacts that plague the shock filter. / Thesis (Master, Mathematics & Statistics) -- Queen's University, 2010-08-25 10:44:12.23

Diese Dissertation beschäftigt sich mit EDP-Konvergenz. Dabei handelt es sich um einen Konvergenzbegriff auf dem Gebiet der verallgemeinerten Gradientensysteme und metrischen Gradientensysteme, der geeignet ist für Gradientenflüsse, die von einem kleinen Parameter abhängen. EDP-Konvergenz liefert einen Algorithmus, der es erlaubt in der Energie und dem Dissipationspotenzial zum Grenzwert überzugehen. Es ist die fundamentale Frage evolutionärer Γ-Konvergenz, wie das Limes-Dissipationspotenzial berechnet werden kann. Das Ziel dieser Arbeit ist es aufzuzeigen, dass EDP-Konvergenz das mikro- und das makroskopische Dissipationspotenzial in einer sinnvollen und eindeutigen Art und Weise in Beziehung setzt. Anhand von drei Beispielen wird der Konvergenzbegriff untersucht: die Diffusionsgleichung auf einem dünnen, dreischichtigen Gebiet, die Poröse-Medien-Gleichung mit einer dünnen Membran und ein Modell mit oszillierender Energie. Es wird die Definition von relaxierter EDP-Konvergenz und EDP-Konvergenz mit Kippung motiviert. EDP-Konvergenz basiert auf dem Prinzip, dass es ein Gleichgewicht zwischen Energie und Dissipation gibt – das Energie-Dissipations-Prinzip (EDP). Mittels Γ-Konvergenz wird sowohl in der Energie, als auch dem totalen Dissipationsfunktional zum Grenzwert übergegangen. Durch die zusätzliche Entkopplung von Zustand und Triebkraft wird die Dissipationslandschaft erkundet und die kinetische Beziehung des Limessystems ermittelt. Das Modell mit oszillierender Energie zeigt die Bedeutung der kinetischen Beziehung – und damit der Kippung – für die Herleitung des Limes-Dissipationspotenzials auf. Die Modelle mit Wasserstein-Dissipation zeigen, dass das Limes-Dissipationspotenzial nicht der naive Grenzwert ist. Insbesondere können klassische Gradientensysteme mit quadratischer Dissipation zu verallgemeinerten Gradientensysteme konvergieren. / In the realm of generalized gradient systems and metric gradient systems we study a notion of convergence suited for gradient flows which depend on a small parameter. This notion is called EDP-convergence. In order to understand the convergence of gradient systems we need an algorithm to derive the limiting energy as well as the limiting dissipation potential. The fundamental question of evolutionary Γ-convergence is how to compute the limit dissipation potential. The aim of this thesis is to show that EDP-convergence connects the microscopic dissipation potential with the macroscopic, i.e. limiting, dissipation potential in a meaningful and unique way. As a proof of concept 3 different examples are presented: (i) the diffusion equation on a thin sandwich-like domain, (ii) the porous medium equation with a thin interface and (iii) a wiggly energy model. We show how the gradient flow concept that is used in this thesis can be used to obtain also gradient flows with respect to the Wasserstein metric. We motivate the definition of relaxed EDP-convergence and EDP- convergence with tilting. EDP-convergence is based upon the principle that there is an energy-dissipation-balance involving the total dissipation functional and the energy difference – the energy-dissipation-principle (EDP). The limit passage, in both the energy and the total dissipation functional, is performed in terms of Γ-convergence. By perturbing the flow as well as the driving force, the dissipation-landscape is explored and a kinetic relation for the limit system can be established. The wiggly energy model demonstrates the importance of the kinetic relation for the construction of the limiting dissipation potential and thus the introduction of tilts. The models with a Wasserstein dissipation show that the limiting dissipation potential is not the naive limit. In particular, classical gradient systems with a quadratic dissipation potential converge to a generalized gradient systems.

In this thesis we study three stochastic partial differential equations (SPDE) that arise as stochastic gradient flows via the fluctuation-dissipation principle. For the first equation we establish a finer regularity statement based on a generalized Taylor expansion which is inspired by the theory of rough paths. The second equation is the thin-film equation with thermal noise which is a singular SPDE. In order to circumvent the issue of dealing with possible renormalization, we discretize the gradient flow structure of the deterministic thin-film equation. Choosing a specific discretization of the metric tensor, we resdiscover a well-known discretization of the thin-film equation introduced by Grün and Rumpf that satisfies a discrete entropy estimate. By proving a stochastic entropy estimate in this discrete setting, we obtain positivity of the scheme in the case of no-slip boundary conditions. Moreover, we analyze the associated rate functional and perform numerical experiments which suggest that the scheme converges. The third equation is the massive $\varphi^4_2$-model on the torus which is also a singular SPDE. In the spirit of Bakry and Émery, we obtain a gradient bound on the Markov semigroup. The proof relies on an $L^2$-estimate for the linearization of the equation. Due to the required renormalization, we use a stopping time argument in order to ensure stochastic integrability of the random constant in the estimate. A postprocessing of this estimate yields an even sharper gradient bound. As a corollary, for large enough mass, we establish a local spectral gap inequality which by ergodicity yields a spectral gap inequality for the $\varphi^4_2$- measure.

Dans cette thèse, nous nous intéressons à la modélisation de mouvements de foules. Nous proposons un modèle microscopique basé sur la théorie des jeux. Chaque individu a une certaine vitesse souhaitée, celle qu'il adopterait en l'absence des autres. Une personne est influencée par certains de ses voisins, pratiquement ceux qu'elle voit devant elle. Une vitesse réelle est considérée comme possible si elle réalise un équilibre de Nash instantané: chaque individu fait son mieux par rapport à un objectif personnel (vitesse souhaitée), en tenant compte du comportement des voisins qui l'influencent. Nous abordons des questions relatives à la modélisation ainsi que les aspects théoriques du problème dans diverses situations, en particulier dans le cas où chaque individu est influencé par tous les autres, et le cas où les relations d'influence entre les individus présentent une structure hiérarchique. Un schéma numérique est développé pour résoudre le problème dans le second cas (modèle hiérarchique) et des simulations numériques sont proposées pour illustrer le comportement du modèle. Les résultats numériques sont confrontés avec des expériences réelles de mouvements de foules pour montrer la capacité du modèle à reproduire certains effets.Nous proposons une version macroscopique du modèle hiérarchique en utilisant les mêmes principes de modélisation au niveau macroscopique, et nous présentons une étude préliminaire des difficultés posées par cette approche.La dernière problématique qu'on aborde dans cette thèse est liée aux cadres flot gradient dans les espaces de Wasserstein aux niveaux continu et discret. Il est connu que l'équation de Fokker-Planck peut s'interpréter comme un flot gradient pour la distance de Wasserstein continue. Nous établissons un lien entre une discrétisation spatiale du type Volume Finis pour l'équation de Fokker-Planck sur une tesselation de Voronoï et les flots gradient sur le réseau sous-jacent, pour une distance de type Wasserstein récemment introduite sur l'espace de mesures portées par les sommets d'un réseaux. / We are interested in the modeling of crowd motion. We propose a microscopic model based on game theoretic principles. Each individual is supposed to have a desired velocity, it is the one he would like to have in the absence of others. We consider that each individual is influenced by some of his neighbors, practically the ones that he sees. A possible actual velocity is an instantaneous Nash equilibrium: each individual does its best with respect to a personal objective (desired velocity), considering the behavior of the neighbors that influence him. We address theoretical and modeling issues in various situations, in particular when each individual is influenced by all the others, and in the case where the influence relations between individuals are hierarchical. We develop a numerical strategy to solve the problem in the second case (hierarchical model) and propose numerical simulations to illustrate the behavior of the model. We confront our numerical results with real experiments and prove the ability of the hierarchical model to reproduce some phenomena.We also propose to write a macroscopic counterpart of the hierarchical model by translating the same modeling principles to the macroscopic level and make the first steps towards writing such model.The last problem tackled in this thesis is related to gradient flow frameworks in the continuous and discrete Wasserstein spaces. It is known that the Fokker-Planck equation can be interpreted as a gradient flow for the continuous Wasserstein distance. We establish a link between some space discretization strategies of the Finite Volume type for the Fokker- Planck equation in general meshes (Voronoï tesselations) and gradient flows on the underlying networks of cells, in the framework of discrete Wasserstein-like distance on graphs recently introduced.

<p dir="ltr">This thesis presents advances in numerical methods for partial differential equations (PDEs) and optimization problems, with a focus on improving efficiency, stability, and accuracy across various applications. We begin by addressing 3D Poisson-type equations, developing a GPU-accelerated spectral-element method that utilizes the tensor product structure to achieve extremely fast performance. This approach enables solving problems with over one billion degrees of freedom in less than one second on modern GPUs, with applications to Schrödinger and Cahn<i>–</i>Hilliard equations demonstrated. Next, we focus on parabolic PDEs, specifically the Cahn<i>–</i>Hilliard equation with dynamical boundary conditions. We propose an efficient energy-stable numerical scheme using a unified framework to handle both Allen<i>–</i>Cahn and Cahn<i>–</i>Hilliard type boundary conditions. The scheme employs a scalar auxiliary variable (SAV) approach to achieve linear, second-order, and unconditionally energy stable properties. Shifting to a machine learning perspective for PDEs, we introduce an unsupervised learning-based numerical method for solving elliptic PDEs. This approach uses deep neural networks to approximate PDE solutions and employs least-squares functionals as loss functions, with a focus on first-order system least-squares formulations. In the realm of optimization, we present an efficient and robust SAV based algorithm for discrete gradient systems. This method modifies the standard SAV approach and incorporates relaxation and adaptive strategies to achieve fast convergence for minimization problems while maintaining unconditional energy stability. Finally, we address optimization in the context of machine learning by developing a structure-guided Gauss<i>–</i>Newton method for shallow ReLU neural network optimization. This approach exploits both the least-squares and neural network structures to create an efficient iterative solver, demonstrating superior performance on challenging function approximation problems. Throughout the thesis, we provide theoretical analysis, efficient numerical implementations, and extensive computational experiments to validate the proposed methods. </p>

Das Thema dieser Dissertation ist die Anwendung von Variationsmethoden auf Evolutionsgleichungen parabolischen und hyperbolischen Typs. Im ersten Teil der Arbeit beschäftigen wir uns mit Reaktions-Diffusions-Systemen, die sich als Gradientensysteme schreiben lassen. Hierbei verstehen wir unter einem Gradientensystem ein Tripel bestehend aus einem Zustandsraum, einem Entropiefunktional und einer Dissipationsmetrik. Wir geben Bedingungen an, die die geodätische Konvexität des Entropiefunktionals sichern. Geodätische Konvexität ist eine wertvolle aber auch starke strukturelle Eigenschaft und schwer zu zeigen. Wir zeigen anhand zahlreicher Beispiele, darunter ein Drift-Diffusions-System, dass dennoch interessante Systeme existieren, die diese Eigenschaft besitzen. Einen weiteren Punkt dieser Arbeit stellt die Anwendung von Gamma-Konvergenz auf Gradientensysteme dar. Wir betrachten hierbei zwei Modellsysteme aus dem Bereich der Mehrskalenprobleme: Erstens, die rigorose Herleitung einer Allen-Cahn-Gleichung mit dynamischen Randbedingungen und zweitens, einer Interface-Bedingung für eine eindimensionale Diffusionsgleichung jeweils aus einem reinen Bulk-System. Im zweiten Teil der Arbeit beschäftigen wir uns mit dem sog. Weighted-Inertia-Dissipation-Energy-Prinzip für Evolutionsgleichungen. Hierbei werden Trajektorien eines Systems als (Grenzwerte von) Minimierer(n) einer parametrisierten Familie von Funktionalen charakterisiert. Dies erlaubt es, Werkzeuge aus der Theorie der Variationsrechung auf Evolutionsprobleme anzuwenden. Wir zeigen, dass Minimierer der WIDE-Funktionale gegen Lösungen des Ausgangsproblems konvergieren. Hierbei betrachten wir getrennt voneinander den Fall des beschränkten und des unbeschränkten Zeitintervalls, die jeweils mit verschiedenen Methoden behandelt werden. / This thesis deals with the application of variational methods to evolution problems governed by partial differential equations. The first part of this work is devoted to systems of reaction-diffusion equations that can be formulated as gradient systems with respect to an entropy functional and a dissipation metric. We provide methods for establishing geodesic convexity of the entropy functional by purely differential methods. Geodesic convexity is beneficial, however, it is a strong structural property of a gradient system that is rather difficult to achieve. Several examples, including a drift-diffusion system, provide a survey on the applicability of the theory. Next, we demonstrate the application of Gamma-convergence, to derive effective limit models for multiscale problems. The crucial point in this investigation is that we rely only on the gradient structure of the systems. We consider two model problems: The rigorous derivation of an Allen-Cahn system with bulk/surface coupling and of an interface condition for a one-dimensional diffusion equation. The second part of this thesis is devoted to the so-called Weighted-Inertia-Dissipation-Energy principle. The WIDE principle is a global-in-time variational principle for evolution equations either of conservative or dissipative type. It relies on the minimization of a specific parameter-dependent family of functionals (WIDE functionals) with minimizers characterizing entire trajectories of the system. We prove that minimizers of the WIDE functional converge, up to subsequences, to weak solutions of the limiting PDE when the parameter tends to zero. The interest for this perspective is that of moving the successful machinery of the Calculus of Variations.

L'objet de cette thèse est d'étendre le cadre théorique et les méthodes numériques du transport optimal à des objets plus généraux que des mesures de probabilité. En premier lieu, nous définissons des modèles de transport optimal entre mesures positives suivant deux approches, interpolation et couplage de mesures, dont nous montrons l'équivalence. De ces modèles découle une généralisation des métriques de Wasserstein. Dans une seconde partie, nous développons des méthodes numériques pour résoudre les deux formulations et étudions en particulier une nouvelle famille d'algorithmes de "scaling", s'appliquant à une grande variété de problèmes. La troisième partie contient des illustrations ainsi que l'étude théorique et numérique, d'un flot de gradient de type Hele-Shaw dans l'espace des mesures. Pour les mesures à valeurs matricielles, nous proposons aussi un modèle de transport optimal qui permet un bon arbitrage entre fidélité géométrique et efficacité algorithmique. / This thesis generalizes optimal transport beyond the classical "balanced" setting of probability distributions. We define unbalanced optimal transport models between nonnegative measures, based either on the notion of interpolation or the notion of coupling of measures. We show relationships between these approaches. One of the outcomes of this framework is a generalization of the p-Wasserstein metrics. Secondly, we build numerical methods to solve interpolation and coupling-based models. We study, in particular, a new family of scaling algorithms that generalize Sinkhorn's algorithm. The third part deals with applications. It contains a theoretical and numerical study of a Hele-Shaw type gradient flow in the space of nonnegative measures. It also adresses the case of measures taking values in the cone of positive semi-definite matrices, for which we introduce a model that achieves a balance between geometrical accuracy and algorithmic efficiency.