With this thesis we bring some new results and improve some existing ones in conjugate duality and some of the areas it is applied in. First we recall the way Lagrange, Fenchel and Fenchel - Lagrange dual problems to a given primal optimization problem can be obtained via perturbations and we present some connections between them. For the Fenchel - Lagrange dual problem we prove strong duality under more general conditions than known so far, while for the Fenchel duality we show that the convexity assumptions on the functions involved can be weakened without altering the conclusion. In order to prove the latter we prove also that some formulae concerning conjugate functions given so far only for convex functions hold also for almost convex, respectively nearly convex functions. After proving that the generalized geometric dual problem can be obtained via perturbations, we show that the geometric duality is a special case of the Fenchel - Lagrange duality and the strong duality can be obtained under weaker conditions than stated in the existing literature. For various problems treated in the literature via geometric duality we show that Fenchel - Lagrange duality is easier to apply, bringing moreover strong duality and optimality conditions under weaker assumptions. The results presented so far are applied also in convex composite optimization and entropy optimization. For the composed convex cone - constrained optimization problem we give strong duality and the related optimality conditions, then we apply these when showing that the formula of the conjugate of the precomposition with a proper convex K - increasing function of a K - convex function on some n - dimensional non - empty convex set X, where K is a k - dimensional non - empty closed convex cone, holds under weaker conditions than known so far. Another field were we apply these results is vector optimization, where we provide a general duality framework based on a more general scalarization that includes as special cases and improves some previous results in the literature. Concerning entropy optimization, we treat first via duality a problem having an entropy - like objective function, from which arise as special cases some problems found in the literature on entropy optimization. Finally, an application of entropy optimization into text classification is presented.

There is a little doubt that, for a decade, risk measurement has become one of the most important topics in finance. Indeed, it is natural to observe such a development, since in the last ten years, huge amounts of financial transactions ended with severe losses due to severe convulsions in financial markets. Value at risk, as the most widely used risk measure, fails to quantify the risk of a position accurately in many situations. For this reason a number of consistent risk measures have been introduced in the literature. The main aim of this study is to present and compare coherent, convex, conditional convex and some other risk measures both in theoretical and practical settings.

This thesis examines the elegant theory of polar and Legendre duality, and its potential use in convex geometry and geometric analysis. It derives a theorem of polar - Legendre duality for all convex bodies, which is captured in a commutative diagram. A geometric flow on a convex body induces a distortion on its polar dual. In general these distortions are not flows defined by local curvature, but in two dimensions they do have similarities to the inverse flows on the original convex bodies. These ideas can be extended to higher dimensions. Polar - Legendre duality can also be used to examine Mahler's Conjecture in convex geometry. The theory presents new insight on the resolved two-dimensional problem, and presents some ideas on new approaches to the still open three dimensional problem.

This dissertation consists of two self-contained studies. The first study, in the domain of stochastic inventory theory, addresses the structure of optimal ordering policies in a periodic review setting. We take multiple sources of a single product to imply an ordering cost function that is nondecreasing, piecewise linear, and convex. Our main contribution is a proof of the optimality of a finite generalized base stock policy under an average cost criterion. Our inventory model is formulated as a Markov decision process with complete observations. Orders are delivered immediately. Excess demand is fully backlogged, and the function describing holding and backlogging costs is convex. All parameters are stationary, and the random demands are independent and identically distributed across periods. The (known) distribution function is subject to mild assumptions along with the holding and backlogging cost function. Our proof uses a vanishing discount approach. We extend our results from a continuous environment to the case where demands and order quantities are integral. The second study is in the area of capacity planning. Our overarching contribution is a relatively simple and fast solution approach for the fleet composition problem faced by a retail distribution firm, focusing on the context of a major beverage distributor. Vehicles to be included in the fleet may be of multiple sizes; we assume that spot transportation capacity will be available to supplement the fleet as needed. We aim to balance the fixed costs of the fleet against exposure to high variable costs due to reliance on spot capacity. We propose a two-stage stochastic linear programming model with fixed recourse. The demand on a particular day in the planning horizon is described by the total quantity to be delivered and the total number of customers to visit. Thus, daily demand throughout the entire planning period is captured by a bivariate probability distribution. We present an algorithm that efficiently generates a "definitive" collection of bases of the recourse program, facilitating rapid computation of the expected cost of a prospective fleet and its gradient. The equivalent convex program may then be solved by a standard gradient projection algorithm.

Constraint handling is difficult in model predictive control (MPC) of linear differential inclusions (LDIs) and linear parameter varying (LPV) systems. The designer is faced with a choice of using conservative bounds that may give poor performance, or accurate ones that require heavy online computation. This thesis presents a framework to achieve a more flexible trade-off between these two extremes by using a state tube, a sequence of parametrised polyhedra that is guaranteed to contain the future state. To define controllers using a tube, one must ensure that the polyhedra are a sub-set of the region defined by constraints. Necessary and sufficient conditions for these subset relations follow from duality theory, and it is possible to apply these conditions to constrain predicted system states and inputs with only a little conservatism. This leads to a general method of MPC design for uncertain-parameter systems. The resulting controllers have strong theoretical properties, can be implemented using standard algorithms and outperform existing techniques. Crucially, the online optimisation used in the controller is a convex problem with a number of constraints and variables that increases only linearly with the length of the prediction horizon. This holds true for both LDI and LPV systems. For the latter it is possible to optimise over a class of gain-scheduled control policies to improve performance, with a similar linear increase in problem size. The framework extends to stochastic LDIs with chance constraints, for which there are efficient suboptimal methods using online sampling. Sample approximations of chance constraint-admissible sets are generally not positively invariant, which motivates the novel concept of ‘sample-admissible' sets with this property to ensure recursive feasibility when using sampling methods. The thesis concludes by introducing a simple, convex alternative to chance-constrained MPC that applies a robust bound to the time average of constraint violations in closed-loop.

Some problems concerning holomorphic continuation of the class of bounded holo­morphic functions from bounded domains in Cn that are domains of holomorphy are solved. A bounded domain of holomorphy Ω in C2 with nonschlicht H°°-envelope of holomorphy is constructed and it is shown that there is a point in D for which Glea­son’s Problem for H°°(Ω) cannot be solved. Furthermore a proof of the existence of a bounded domain of holomorphy in C2 for which the volume of the H°°-envelope of holomorphy is infinite is given. The idea of the proof is to put a family of so-called ”Sibony domains” into the unit bidisk by a packing procedure and patch them together by thin neighbourhoods of suitably chosen curves. If H°°(Ω) is the Banach algebra of bounded holomorphic functions on a bounded domain Ω in Cn and if p is a point in Ω, then the following problem is known as Gleason’s Problem for Hoo(Ω) : Is the maximal ideal in H°°(Ω) consisting of functions vanishing at p generated by (z1 -p1) , ... ,   (zn - pn) ? A sufficient condition for solving Gleason’s Problem for 77°° (Ω) for all points in Ω is given. In particular, this condition is fulfilled by a convex domain Ω with Lip1+e boundary (0 < e < 1) and thus generalizes a theorem of S.L.Leibenson. It is also proved that Gleason’s Problem can be solved for all points in certain unions of two polydisks in C2. One of the ideas in the methods of proof is integration along specific polygonal lines. Certain properties of some open sets defined by global plurisubharmonic func­tions in Cn are studied. More precisely, the sets Du = {z e Cn : u(z) < 0} and Eh = {{z,w) e Cn X C : h(z,w) < 1} are considered where ti is a plurisubharmonic function of minimal growth and h≠0 is a non-negative homogeneous plurisubharmonic function. (That is, the functions u and h belong to the classes L(Cn) and H+(Cn x C) respectively.) It is examined how the fact that Eh and the connected components of Du are H°°-domains of holomorphy is related to the structure of the set of disconti­nuity points of the global defining functions and to polynomial convexity. Moreover it is studied whether these notions are preserved under a certain bijective mapping from L(Cn) to H+(Cn x C). Two counterexamples are given which show that polynomial convexity is not preserved under this bijection. It is also proved, for example, that if Du is bounded and if the set of discontinuity points of u is pluripolar then Du is of type H°°. A survey paper on general properties of envelopes of holomorphy is included. In particular, the paper treats aspects of the theory for the bounded holomorphic functions. The results for the bounded holomorphic functions are compared with the corresponding ones for the holomorphic functions. / digitalisering@umu.se

Let K be a compact subset of Cn . The polynomially convex hull of K is defined as The compact set K is said to be polynomially convex if = K. A closed subset is said to be locally polynomially convex at if there exists a closed ball centred at z such that is polynomially convex. The aim of this thesis is to derive easily checkable conditions to detect polynomial convexity in certain classes of sets in This thesis begins with the basic question: Let S1 and S2 be two smooth, totally real surfaces in C2 that contain the origin. If the union of their tangent planes is locally polynomially convex at the origin, then is locally polynomially convex at the origin? If then it is a folk result that the answer is, “Yes.” We discuss an obstruction to the presumed proof, and use a different approach to provide a proof. When dimR it turns out that the positioning of the complexification of controls the outcome in many situations. In general, however, local polynomial convexity of also depends on the degeneracy of the contact of T0Sj with We establish a result showing this. Next, we consider a generalization of Weinstock’s theorem for more than two totally real planes in C2 . Using a characterization, recently found by Florentino, for simultaneous triangularizability over R of real matrices, we present a sufficient condition for local polynomial convexity at of union of finitely many totally real planes is C2 . The next result is motivated by an approximation theorem of Axler and Shields, which says that the uniform algebra on the closed unit disc generated by z and h — where h is a nowhereholomorphic harmonic function on D that is continuous up to ∂D — equals . The abstract tools used by Axler and Shields make harmonicity of h an essential condition for their result. We use the concepts of plurisubharmonicity and polynomial convexity to show that, in fact, the same conclusion is reached if h is replaced by h+ R, where R is a nonharmonic perturbation whose Laplacian is “small” in a certain sense. Ideas developed for the latter result, especially the role of plurisubharmonicity, lead us to our final result: a characterization for compact patches of smooth, totallyreal graphs in to be polynomially convex.

The following problem has been known for its beauty and elementary character. The Erd˝os Szekeres problem[7]: For any integer k ≥ 3, determine if there exists a smallest positive integer N(k) such that any set of atleast N(k) points in general position in the plane(i.e no three points are in a line) contains k points that are the vertices of a convex k-gon. The finiteness of (k)is proved by Erd˝os and Szekeres using Ramsey theory[7]. In 1978, Erd˝os [6] raised a similar question on empty convex k-gon (convex k-gon without out any interior points) and it has been extensively studied[18]. Several other variants like the convex k-gon with specified number interior points[2] and the chromatic variant[5] have been well studied. In this thesis, we introduce the following two player game variant of the Erd˝os Szekeres problem: Consider a two player game where each player places a point in the plane such that the point set formed will not contain a convex k-gon. The game will end when a convex k-gon is formed and the player who placed the last point loses the game. In our thesis we show a winning strategy forplayer2inthe convex5-gongame and the empty convex5-gongame and argue that the game always ends in the 9th step. We also give an alternative proof for the statement that any point set containing 10 or more points contains an empty convex 5-gon.

Ce travail de thèse porte principalement sur deux problématiques différentes mais qui ont pour point commun, la contribution à la modélisation et à la gestion du risque en actuariat. Dans le premier thème de recherche abordé dans cette thèse, on s'intéresse à la modélisation de la dépendance en assurance et en particulier, on propose une extension des modèles à facteurs communs qui sont utilisés en assurance. Dans le deuxième thème de recherche, on considère les distributions discrètes décroissantes et on s'intéresse à l'étude de l'effet de l'ajout de la contrainte de convexité sur les extrema convexes. Des applications en liaison avec la théorie de la ruine motivent notre intérêt pour ce sujet. Dans la première partie de la thèse, on considère un modèle de risque en temps discret dans lequel les variables aléatoires sont dépendantes mais conditionnellement indépendantes par rapport à un facteur commun. Dans ce cadre de dépendance on introduit un nouveau concept pour la modélisation de la dépendance temporelle entre les risques d'un portefeuille d'assurance. En effet, notre modélisation inclut des processus de mémoire non bornée. Plus précisément, le conditionnement se fait par rapport à un vecteur aléatoire de longueur variable au cours du temps. Sous des conditions de mélange du facteur et d'une structure de mélange conditionnel, nous avons obtenu des propriétés de mélanges pour les processus non conditionnels. Avec ces résultats on peut obtenir des propriétés asymptotiques intéressantes. On note que dans notre étude asymptotique c'est plutôt le temps qui tend vers l'infini que le nombre de risques. On donne des résultats asymptotiques pour le processus agrégé, ce qui permet de donner une approximation du risque d'une compagnie d'assurance lorsque le temps tend vers l'infini. La deuxième partie de la thèse porte sur l'effet de la contrainte de convexité sur les extrema convexes dans la classe des distributions discrètes dont les fonctions de masse de probabilité (f.m.p.) sont décroissantes sur un support fini. Les extrema convexes dans cette classe de distributions sont bien connus. Notre but est de souligner comment les contraintes de forme supplémentaires de type convexité modifient ces extrema. Deux cas sont considérés : la f.m.p. est globalement convexe sur N et la f.m.p. est convexe seulement à partir d'un point positif donné. Les extrema convexes correspondants sont calculés en utilisant de simples propriétés de croisement entre deux distributions. Plusieurs illustrations en théorie de la ruine sont présentées / In this thesis we focus on two different problems which have as common point the contribution to the modeling and to the risk management in insurance. In the first research theme, we are interested by the modeling of the dependence in insurance. In particular we propose an extension to model with common factor. In the second research theme we consider the class of nonincreasing discrete distributions and we are interested in studying the effect of additional constraint of convexity on the convex extrema. Some applications in ruin theory motivate our interest to this subject. The first part of this thesis is concerned with factor models for the modeling of the dependency in insurance. An interesting property of these models is that the random variables are conditionally independent with respect to a factor. We propose a new model in which the conditioning is with respect to the entire memory of the factor. In this case we give some mixing properties of risk process under conditions related to the mixing properties of the factor process and to the conditional mixing risk process. The law of the sum of random variables has a great interest in actuarial science. Therefore we give some conditions under which the law of the aggregated process converges to a normal distribution. In the second part of the thesis we consider the class of discrete distributions whose probability mass functions (p.m.f.) are nonincreasing on a finite support. Convex extrema in that class of distributions are well-known. Our purpose is to point out how additional shape constraints of convexity type modify these extrema. Two cases are considered : the p.m.f. is globally convex on N or it is convex only from a given positive point. The corresponding convex extrema are derived by using a simple crossing property between two distributions. Several applications to some ruin problems are presented for illustration

De nombreux problèmes de la vie réelle sont exprimés sous la forme de décisions à prendre à l’aide de l’information accessible dans le but d’atteindre certains objectifs. La programmation numérique a prouvé être un outil efficace pour modéliser et résoudre une grande variété de problèmes de ce type. Cependant, de nombreux problèmes en apparence faciles sont encore durs à résoudre. Et même des problèmes faciles de programmation linéaire deviennent durs avec l’incertitude de l’information disponible. Motivés par un problème de télécommunication où l’on doit associer des machines virtuelles à des serveurs tout en minimisant les coûts, nous avons employé plusieurs outils de programmation mathématique dans le but de résoudre efficacement le problème, et développé de nouveaux outils pour des problèmes plus généraux. Dans l’ensemble, résumons les principaux résultats de cette thèse comme suit. Une formulation exacte et plusieurs reformulations pour le problème d’affectation de machines virtuelles dans le cloud sont données. Nous utilisons plusieurs inégalités valides pour renforcer la formulation exacte, accélérant ainsi l’algorithme de résolution de manière significative. Nous donnons en outre un résultat géométrique sur la qualité de la borne lagrangienne montrant qu’elle est généralement beaucoup plus forte que la borne de la relaxation continue. Une hiérarchie de relaxation est également proposée en considérant une séquence de couverture de l’ensemble de la demande. Ensuite, nous introduisons une nouvelle formulation induite par les symétries du problème. Cette formulation permet de réduire considérablement le nombre de termes bilinéaires dans le modèle, et comme prévu, semble plus efficace que les modèles précédents. Deux approches sont développées pour la construction d’enveloppes convexes et concaves pour l’optimisation bilinéaire sur un hypercube. Nous établissons plusieurs connexions théoriques entre différentes techniques et nous discutons d’autres extensions possibles. Nous montrons que deux variantes de formulations pour approcher l’enveloppe convexe des fonctions bilinéaires sont équivalentes. Nous introduisons un nouveau paradigme sur les problèmes linéaires généraux avec des paramètres incertains. Nous proposons une hiérarchie convergente de problèmes d’optimisation robuste – approche robuste multipolaire, qui généralise les notions de robustesse statique, de robustesse d’affinement ajustable, et de robustesse entièrement ajustable. En outre, nous montrons que l’approche multipolaire peut générer une séquence de bornes supérieures et une séquence de bornes inférieures en même temps et les deux séquences convergent vers la valeur robuste des FARC sous certaines hypothèses modérées / Many real life problems are characterized by making decisions with current information to achieve certain objectives. Mathematical programming has been developed as a successful tool to model and solve a wide range of such problems. However, many seemingly easy problems remain challenging. And some easy problems such as linear programs can be difficult in the face of uncertainty. Motivated by a telecommunication problem where assignment decisions have to be made such that the cloud virtual machines are assigned to servers in a minimum-cost way, we employ several mathematical programming tools to solve the problem efficiently and develop new tools for general theoretical problems. In brief, our work can be summarized as follows. We provide an exact formulation and several reformulations on the cloud virtual machine assignment problem. Then several valid inequalities are used to strengthen the exact formulation, thereby accelerating the solution procedure significantly. In addition, an effective Lagrangian decomposition is proposed. We show that, the bounds providedby the proposed Lagrangian decomposition is strong, both theoretically and numerically. Finally, a symmetry-induced model is proposed which may reduce a large number of bilinear terms in some special cases. Motivated by the virtual machine assignment problem, we also investigate a couple of general methods on the approximation of convex and concave envelopes for bilinear optimization over a hypercube. We establish several theoretical connections between different techniques and prove the equivalence of two seeming different relaxed formulations. An interesting research direction is also discussed. To address issues of uncertainty, a novel paradigm on general linear problems with uncertain parameters are proposed. This paradigm, termed as multipolar robust optimization, generalizes notions of static robustness, affinely adjustable robustness, fully adjustable robustness and fills the gaps in-between. As consequences of this new paradigms, several known results are implied. Further, we prove that the multipolar approach can generate a sequence of upper bounds and a sequence of lower bounds at the same time and both sequences converge to the robust value of fully adjustable robust counterpart under some mild assumptions