Les systèmes intégrables toriques sont des systèmes intégrables dont toutes les composantes de l'application moment sont périodiques de même période. Il s'agit donc de variétés symplectiques munies d'actions Hamiltoniennes de tores. Au début des années 80, Atiyah-Guillemin-Sternberg ont démontré que l'image de l'application moment était un polytope convexe à face rationnelles. Peu de temps après, Delzant a démontré que dans le cas intégrable qui nous intéresse, ce polytope caractérisait entièrement le système : la variété symplectique comme l'action du tore. Le champs d'étude s'est ensuite élargi aux systèmes dits semi-toriques. Ce sont des systèmes intégrables dont toutes les composantes de l'application moment sauf une sont périodiques de même période. En outre, pour simplifier l'étude de ces systèmes, on demande que tous les points critiques du systèmes soient non-dégénérés, et sans composante hyperbolique pour la hessienne. En revanche les points critiques des systèmes semi-toriques peuvent comporter des composantes dites "foyer-foyer". Celles-ci ont une dynamique plus riche que les singularités elliptiques, mais conservent certaines propriétés qui rendent leur analyse plus aisée que les singularités hyperboliques. San Vu-Ngoc et Alvaro Pelayo ont réussi à étendre pour ces systèmes semi-toriques les résultats d'Atiyah-Guillemin-Sternberg et Delzant en dimension 2. L'objectif de cette thèse est de proposer une extension de ces résultats en dimension quelconque, à commencer par la dimension 3. Les techniques utilisées relèvent de l'analyse comme de la géométrie symplectique, ainsi que de la théorie de Morse dans des espaces différentiels stratifiés. / Semi-toric integrable systems are integrable systems whose every component of the moment map are periodic of the same period. They are symplectic manifolds endowed with a Hamiltonian torus actions. At the beginning of the 80's, Atiyah-Guillemin-Sternberg proved that the image of the moment map was a polytope with rational faces. A bit after that, Delzant showed that in the integrable case that matters to us, this polytope characterized entirely the system, that is, the symplectic manifold as well as the torus action. Next, field of study widened to semi-toric systems. They are integrable systems whose all components except one are periodic with the same period. Moreover, to simplify their study, we ask that these systems have only non-degenerate critical points without hyperbolic components. On the other hand, critical points of semi-toric systems can have so-called ''focus-focus'' components. They have a richer dynamic than elliptic singularities, but it retains some properties that makes them easier to study than hyperbolic singularities. San Vu-Ngoc and Alvaro Pelayo have managed to extend to these semi-toric systems the results of Atiyah-Guillemin-Sternberg and Delzant in dimension 2. The objective of this thesis is to propose an extension of these results to any dimension, starting with dimension 3. Techniques involved are analysis as well as symplectic geometry, and Morse theory in stratified differential spaces.

A spectral graph realization is an embedding of a finite simple graph into Euclidean space that is constructed from the eigenvalues and eigenvectors of the graph's adjacency matrix. It has previously been observed that some polytopes can be reconstructed from their edge-graphs by taking the convex hull of a spectral realization of this edge-graph. These polytopes, which we shall call spectral polytopes, have remarkable rigidity and symmetry properties and are a source for many open questions. In this thesis we aim to further the understanding of this phenomenon by exploring the geometric and combinatorial properties of spectral polytopes on several levels. One of our central questions is whether already weak forms of symmetry can be a sufficient reason for a polytope to be spectral. To answer this, we derive a geometric criterion for the identification of spectral polytopes and apply it to prove that indeed all polytopes of combined vertex- and edge-transitivity are spectral, admit a unique reconstruction from the edge-graph and realize all the symmetries of this edge-graph. We explore the same questions for graph realizations and find that realizations of combined vertex- and edge-transitivity are not necessarily spectral. Instead we show that we require a stronger form of symmetry, called distance-transitivity. Motivated by these findings we take a closer look at the class of edge-transitive polytopes, for which no classification is known. We state a conjecture for a potential classification and provide complete classifications for several sub-classes, such as distance-transitive polytopes and edge-transitive polytopes that are not vertex-transitive. In particular, we show that the latter class contains only polytopes of dimension d <= 3. As a side result we obtain the complete classification of the vertex-transitive zonotopes and a new characterization for root systems.

By the name HALFSPACE SYSTEMS, this dissertation refers to systems whose dynamics are modeled by linear constraints of the form Exk+1 <= Fxk + Buk (where E, F 2 andlt;mn, B 2 andlt;mp). This dissertation explores the concepts of BOUNDEDNESS, STABILITY, IRREDUNDANCY, and MAINTAINABILITY (which is the same as REACHABILITY OF A TARGET TUBE) that are related to the control of halfspace systems. Given that a halfspace system is bounded, and that a given static target tube is reachable for this system, this dissertation presents algorithms to MAINTAIN the system in this target tube. A DIFFERENCE INCLUSION has the form xk+1 = Axk + Buuk, where xk, xk+1 2 andlt;n, uk 2 andlt;p, A 2 andlt;nn, Bu 2 andlt;np, Ai 2 andlt;nn, Bj 2 andlt;np, and A and Bu belong to the convex hulls of (A1,A2, . . . ,Aq) and (B1, B2, . . . , Br) respectively. This dissertation investigates the possibility that halfspace systems have equivalent difference inclusion representation for the case of uk = 0. An affirmitive result in this direction may make it possible to apply to halfspace systems the control theory that exists for difference inclusions.

We study the relationship between geometric properties of toric varieties and combinatorial properties of the corresponding lattice polytopes. In particular, we give a bound for a very ample lattice polytope to be k-normal. Equivalently, we give a new combinatorial bound for the Castelnuovo-Mumford regularity of normal projective toric varieties. We also give a new combinatorial proof for a special case of Reider's Theorem for smooth toric surfaces.

Simplices are the "simplest" examples of polytopes, and yet they exhibit much of the rich and subtle combinatorics and commutative algebra of their more general cousins. In this way they are sufficiently complicated --- insights gained from their study can inform broader research in Ehrhart theory and associated fields. In this dissertation we consider two previously unstudied properties of lattice simplices; one algebraic and one combinatorial. The first is the Poincar\'e series of the associated semigroup algebra, which is substantially more complicated than the Hilbert series of that same algebra. The second is the partial ordering of the elements of the fundamental parallelepiped associated to the simplex. We conclude with a proof-of-concept for using machine learning techniques in algebraic combinatorics. Specifically, we attempt to model the integer decomposition property of a family of lattice simplices using a neural network.

.

This thesis is based on three papers on selected topics in Asymptotic Geometric Analysis. The first paper is about the volume of high-dimensional random polytopes; in particular, on polytopes generated by Gaussian random vectors. We consider the question of how many random vertices (or facets) should be sampled in order for such a polytope to capture significant volume. Various criteria for what exactly it means to capture significant volume are discussed. We also study similar problems for random polytopes generated by points on the Euclidean sphere. The second paper is about volume distribution in convex bodies. The first main result is about convex bodies that are (i) symmetric with respect to each of the coordinate hyperplanes and (ii) in isotropic position. We prove that most linear functionals acting on such bodies exhibit super-Gaussian tail-decay. Using known facts about the mean-width of such bodies, we then deduce strong lower bounds for the volume of certain caps. We also prove a converse statement. Namely, if an arbitrary isotropic convex body (not necessarily satisfying the symmetry assumption (i)) exhibits similar cap-behavior, then one can bound its mean-width. The third paper is about random polytopes generated by sampling points according to multiple log-concave probability measures. We prove related estimates for random determinants and give applications to several geometric inequalities; these include estimates on the volume-radius of random zonotopes and Hadamard's inequality for random matrices. / Mathematics

Reproduction de : Thèse de doctorat : Mathématiques appliquées : Grenoble 1 : 1995. / Titre provenant de l'écran-titre. Bibliogr. p. 169-174.

This thesis focuses on applications of classical tools from probability theory and convex analysis such as limit theorems to problems in theoretical computer science, specifically to pseudorandomness and learning theory. At first look, limit theorems, pseudorandomness and learning theory appear to be disparate subjects. However, as it has now become apparent, there's a strong connection between these questions through a third more abstract question: what do random objects look like. This connection is best illustrated by the study of the spectrum of Boolean functions which directly or indirectly played an important role in a plethora of results in complexity theory. The current thesis aims to take this program further by drawing on a variety of fundamental tools, both classical and new, in probability theory and analytic geometry. Our research contributions broadly fall into three categories. Probability Theory: The central limit theorem is one of the most important results in all of probability and richly studied topic. Motivated by questions in pseudorandomness and learning theory we obtain two new limit theorems or invariance principles. The proofs of these new results in probability, of interest on their own, have a computer science flavor and fall under the niche category of techniques from theoretical computer science with applications in pure mathematics. Pseudorandomness: Derandomizing natural complexity classes is a fundamental problem in complexity theory, with several applications outside complexity theory. Our work addresses such derandomization questions for natural and basic geometric concept classes such as halfspaces, polynomial threshold functions (PTFs) and polytopes. We develop a reasonably generic framework for obtaining pseudorandom generators (PRGs) from invariance principles and suitably apply the framework to old and new invariance principles to obtain the best known PRGs for these complexity classes. Learning Theory: Learning theory aims to understand what functions can be learned efficiently from examples. As developed in the seminal work of Linial, Mansour and Nisan (1994) and strengthened by several follow-up works, we now know strong connections between learning a class of functions and how sensitive to noise, as quantified by average sensitivity and noise sensitivity, the functions are. Besides their applications in learning, bounding the average and noise sensitivity has applications in hardness of approximation, voting theory, quantum computing and more. Here we address the question of bounding the sensitivity of polynomial threshold functions and intersections of halfspaces and obtain the best known results for these concept classes.

No description available.