• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 12
  • 2
  • 2
  • 1
  • Tagged with
  • 20
  • 20
  • 5
  • 5
  • 5
  • 4
  • 4
  • 4
  • 4
  • 4
  • 3
  • 3
  • 3
  • 3
  • 3
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
11

Techniques avancées d'apprentissage automatique basées sur la programmation DC et DCA / Advanced machine learning techniques based on DC programming and DCA

Ho, Vinh Thanh 08 December 2017 (has links)
Dans cette thèse, nous développons certaines techniques avancées d'apprentissage automatique dans le cadre de l'apprentissage en ligne et de l'apprentissage par renforcement (« reinforcement learning » en anglais -- RL). L'épine dorsale de nos approches est la programmation DC (Difference of Convex functions) et DCA (DC Algorithm), et leur version en ligne, qui sont reconnues comme de outils puissants d'optimisation non convexe, non différentiable. Cette thèse se compose de deux parties : la première partie étudie certaines techniques d'apprentissage automatique en mode en ligne et la deuxième partie concerne le RL en mode batch et mode en ligne. La première partie comprend deux chapitres correspondant à la classification en ligne (chapitre 2) et la prédiction avec des conseils d'experts (chapitre 3). Ces deux chapitres mentionnent une approche unifiée d'approximation DC pour différents problèmes d'optimisation en ligne dont les fonctions objectives sont des fonctions de perte 0-1. Nous étudions comment développer des algorithmes DCA en ligne efficaces en termes d'aspects théoriques et computationnels. La deuxième partie se compose de quatre chapitres (chapitres 4, 5, 6, 7). Après une brève introduction du RL et ses travaux connexes au chapitre 4, le chapitre 5 vise à fournir des techniques efficaces du RL en mode batch basées sur la programmation DC et DCA. Nous considérons quatre différentes formulations d'optimisation DC en RL pour lesquelles des algorithmes correspondants basés sur DCA sont développés. Nous traitons les problèmes clés de DCA et montrons l'efficacité de ces algorithmes au moyen de diverses expériences. En poursuivant cette étude, au chapitre 6, nous développons les techniques du RL basées sur DCA en mode en ligne et proposons leurs versions alternatives. Comme application, nous abordons le problème du plus court chemin stochastique (« stochastic shortest path » en anglais -- SSP) au chapitre 7. Nous étudions une classe particulière de problèmes de SSP qui peut être reformulée comme une formulation de minimisation de cardinalité et une formulation du RL. La première formulation implique la norme zéro et les variables binaires. Nous proposons un algorithme basé sur DCA en exploitant une approche d'approximation DC de la norme zéro et une technique de pénalité exacte pour les variables binaires. Pour la deuxième formulation, nous utilisons un algorithme batch RL basé sur DCA. Tous les algorithmes proposés sont testés sur des réseaux routiers artificiels / In this dissertation, we develop some advanced machine learning techniques in the framework of online learning and reinforcement learning (RL). The backbones of our approaches are DC (Difference of Convex functions) programming and DCA (DC Algorithm), and their online version that are best known as powerful nonsmooth, nonconvex optimization tools. This dissertation is composed of two parts: the first part studies some online machine learning techniques and the second part concerns RL in both batch and online modes. The first part includes two chapters corresponding to online classification (Chapter 2) and prediction with expert advice (Chapter 3). These two chapters mention a unified DC approximation approach to different online learning algorithms where the observed objective functions are 0-1 loss functions. We thoroughly study how to develop efficient online DCA algorithms in terms of theoretical and computational aspects. The second part consists of four chapters (Chapters 4, 5, 6, 7). After a brief introduction of RL and its related works in Chapter 4, Chapter 5 aims to provide effective RL techniques in batch mode based on DC programming and DCA. In particular, we first consider four different DC optimization formulations for which corresponding attractive DCA-based algorithms are developed, then carefully address the key issues of DCA, and finally, show the computational efficiency of these algorithms through various experiments. Continuing this study, in Chapter 6 we develop DCA-based RL techniques in online mode and propose their alternating versions. As an application, we tackle the stochastic shortest path (SSP) problem in Chapter 7. Especially, a particular class of SSP problems can be reformulated in two directions as a cardinality minimization formulation and an RL formulation. Firstly, the cardinality formulation involves the zero-norm in objective and the binary variables. We propose a DCA-based algorithm by exploiting a DC approximation approach for the zero-norm and an exact penalty technique for the binary variables. Secondly, we make use of the aforementioned DCA-based batch RL algorithm. All proposed algorithms are tested on some artificial road networks
12

Descent dynamical systems and algorithms for tame optimization, and multi-objective problems / Systèmes dynamiques de descente et algorithmes pour l'optimisation modérée, et les problèmes multi-objectif

Garrigos, Guillaume 02 November 2015 (has links)
Dans une première partie, nous nous intéressons aux systèmes dynamiques gradients gouvernés par des fonctions non lisses, mais aussi non convexes, satisfaisant l'inégalité de Kurdyka-Lojasiewicz. Après avoir obtenu quelques résultats préliminaires pour la dynamique de la plus grande pente continue, nous étudions un algorithme de descente général. Nous prouvons, sous une hypothèse de compacité, que tout suite générée par ce schéma général converge vers un point critique de la fonction. Nous obtenons aussi de nouveaux résultats sur la vitesse de convergence, tant pour les valeurs que pour les itérés. Ce schéma général couvre en particulier des versions parallélisées de la méthode forward-backward, autorisant une métrique variable et des erreurs relatives. Cela nous permet par exemple de proposer une version non convexe non lisse de l'algorithme Levenberg-Marquardt. Enfin, nous proposons quelques applications de ces algorithmes aux problèmes de faisabilité, et aux problèmes inverses. Dans une seconde partie, cette thèse développe une dynamique de descente associée à des problèmes d'optimisation vectoriels sous contrainte. Pour cela, nous adaptons la dynamique de la plus grande pente usuelle aux fonctions à valeurs dans un espace ordonné par un cône convexe fermé solide. Cette dynamique peut être vue comme l'analogue continu de nombreux algorithmes développés ces dernières années. Nous avons un intérêt particulier pour les problèmes de décision multi-objectifs, pour lesquels cette dynamique de descente fait décroitre toutes les fonctions objectif au cours du temps. Nous prouvons l'existence de trajectoires pour cette dynamique continue, ainsi que leur convergence vers des points faiblement efficients. Finalement, nous explorons une nouvelle dynamique inertielle pour les problèmes multi-objectif, avec l'ambition de développer des méthodes rapides convergeant vers des équilibres de Pareto. / In a first part, we focus on gradient dynamical systems governed by non-smooth but also non-convex functions, satisfying the so-called Kurdyka-Lojasiewicz inequality.After obtaining preliminary results for a continuous steepest descent dynamic, we study a general descent algorithm. We prove, under a compactness assumption, that any sequence generated by this general scheme converges to a critical point of the function.We also obtain new convergence rates both for the values and the iterates. The analysis covers alternating versions of the forward-backward method, with variable metric and relative errors. As an example, a non-smooth and non-convex version of the Levenberg-Marquardt algorithm is detailed.Applications to non-convex feasibility problems, and to sparse inverse problems are discussed.In a second part, the thesis explores descent dynamics associated to constrained vector optimization problems. For this, we adapt the classic steepest descent dynamic to functions with values in a vector space ordered by a solid closed convex cone. It can be seen as the continuous analogue of various descent algorithms developed in the last years.We have a particular interest for multi-objective decision problems, for which the dynamic make decrease all the objective functions along time.We prove the existence of trajectories for this continuous dynamic, and show their convergence to weak efficient points.Then, we explore an inertial dynamic for multi-objective problems, with the aim to provide fast methods converging to Pareto points.
13

Algorithms for structured nonconvex optimization: theory and practice

Nguyen, Hieu Thao 17 October 2017 (has links)
No description available.
14

Optimization techniques for radio resource management in wireless communication networks

Weeraddana, P. C. (Pradeep Chathuranga) 22 November 2011 (has links)
Abstract The application of optimization techniques for resource management in wireless communication networks is considered in this thesis. It is understood that a wide variety of resource management problems of recent interest, including power/rate control, link scheduling, cross-layer control, network utility maximization, beamformer design of multiple-input multiple-output networks, and many others are directly or indirectly reliant on the general weighted sum-rate maximization (WSRMax) problem. Thus, in this dissertation a greater emphasis is placed on the WSRMax problem, which is known to be NP-hard. A general method, based on the branch and bound technique, is developed, which solves globally the nonconvex WSRMax problem with an optimality certificate. Efficient analytic bounding techniques are derived as well. More broadly, the proposed method is not restricted to WSRMax. It can also be used to maximize any system performance metric, which is Lipschitz continuous and increasing on signal-to-interference-plus-noise ratio. The method can be used to find the optimum performance of any network design method, which relies on WSRMax, and therefore it is also useful for evaluating the performance loss encountered by any heuristic algorithm. The considered link-interference model is general enough to accommodate a wide range of network topologies with various node capabilities, such as singlepacket transmission, multipacket transmission, simultaneous transmission and reception, and many others. Since global methods become slow in large-scale problems, fast local optimization methods for the WSRMax problem are also developed. First, a general multicommodity, multichannel wireless multihop network where all receivers perform singleuser detection is considered. Algorithms based on homotopy methods and complementary geometric programming are developed for WSRMax. They are able to exploit efficiently the available multichannel diversity. The proposed algorithm, based on homotopy methods, handles efficiently the self interference problem that arises when a node transmits and receives simultaneously in the same frequency band. This is very important, since the use of supplementary combinatorial constraints to prevent simultaneous transmissions and receptions of any node is circumvented. In addition, the algorithm together with the considered interference model, provide a mechanism for evaluating the gains when the network nodes employ self interference cancelation techniques with different degrees of accuracy. Next, a similar multicommodity wireless multihop network is considered, but all receivers perform multiuser detection. Solutions for the WSRMax problem are obtained by imposing additional constraints, such as that only one node can transmit to others at a time or that only one node can receive from others at a time. The WSRMax problem of downlink OFDMA systems is also considered. A fast algorithm based on primal decomposition techniques is developed to jointly optimize the multiuser subcarrier assignment and power allocation to maximize the weighted sum-rate (WSR). Numerical results show that the proposed algorithm converges faster than Lagrange relaxation based methods. Finally, a distributed algorithm for WSRMax is derived in multiple-input single-output multicell downlink systems. The proposed method is based on classical primal decomposition methods and subgradient methods. It does not rely on zero forcing beamforming or high signal-to-interference-plus-noise ratio approximation like many other distributed variants. The algorithm essentially involves coordinating many local subproblems (one for each base station) to resolve the inter-cell interference such that the WSR is maximized. The numerical results show that significant gains can be achieved by only a small amount of message passing between the coordinating base stations, though the global optimality of the solution cannot be guaranteed. / Tiivistelmä Tässä työssä tutkitaan optimointimenetelmien käyttöä resurssienhallintaan langattomissa tiedonsiirtoverkoissa. Monet ajankohtaiset resurssienhallintaongelmat, kuten esimerkiksi tehonsäätö, datanopeuden säätö, radiolinkkien ajastus, protokollakerrosten välinen optimointi, verkon hyötyfunktion maksimointi ja keilanmuodostus moniantenniverkoissa, liittyvät joko suoraan tai epäsuorasti painotetun summadatanopeuden maksimointiongelmaan (weighted sum-rate maximization, WSRMax). Tästä syystä tämä työ keskittyy erityisesti WSRMax-ongelmaan, joka on tunnetusti NP-kova. Työssä kehitetään yleinen branch and bound -tekniikkaan perustuva menetelmä, joka ratkaisee epäkonveksin WSRMax-ongelman globaalisti ja tuottaa todistuksen ratkaisun optimaalisuudesta. Työssä johdetaan myös tehokkaita analyyttisiä suorituskykyrajojen laskentatekniikoita. Ehdotetun menetelmän käyttö ei rajoitu vain WSRMax-ongelmaan, vaan sitä voidaan soveltaa minkä tahansa suorituskykymetriikan maksimointiin, kunhan se on Lipschitz-jatkuva ja kasvava signaali-häiriö-plus-kohinasuhteen funktiona. Menetelmää voidaan käyttää minkä tahansa WSRMax-ongelmaan perustuvan verkkosuunnittelumenetelmän optimaalisen suorituskyvyn määrittämiseen, ja siksi sitä voidaan hyödyntää myös minkä tahansa heuristisen algoritmin aiheuttaman suorituskykytappion arvioimiseen. Tutkittava linkki-häiriömalli on riittävän yleinen monien erilaisten verkkotopologioiden ja verkkosolmujen kyvykkyyksien mallintamiseen, kuten esimerkiksi yhden tai useamman datapaketin siirtoon sekä yhtäaikaiseen lähetykseen ja vastaanottoon. Koska globaalit menetelmät ovat hitaita suurien ongelmien ratkaisussa, työssä kehitetään WSRMax-ongelmalle myös nopeita paikallisia optimointimenetelmiä. Ensiksi käsitellään yleistä useaa eri yhteyspalvelua tukevaa monikanavaista langatonta monihyppyverkkoa, jossa kaikki vastaanottimet suorittavat yhden käyttäjän ilmaisun, ja kehitetään algoritmeja, joiden perustana ovat homotopiamenetelmät ja komplementaarinen geometrinen optimointi. Ne hyödyntävät tehokkaasti saatavilla olevan monikanavadiversiteetin. Esitetty homotopiamenetelmiin perustuva algoritmi käsittelee tehokkaasti itsehäiriöongelman, joka syntyy, kun laite lähettää ja vastaanottaa samanaikaisesti samalla taajuuskaistalla. Tämä on tärkeää, koska näin voidaan välttää lisäehtojen käyttö yhtäaikaisen lähetyksen ja vastaanoton estämiseksi. Lisäksi algoritmi yhdessä tutkittavan häiriömallin kanssa auttaa arvioimaan, paljonko etua saadaan, kun laitteet käyttävät itsehäiriön poistomenetelmiä erilaisilla tarkkuuksilla. Seuraavaksi tutkitaan vastaavaa langatonta monihyppyverkkoa, jossa kaikki vastaanottimet suorittavat monen käyttäjän ilmaisun. Ratkaisuja WSRMax-ongelmalle saadaan asettamalla lisäehtoja, kuten että vain yksi lähetin kerrallaan voi lähettää tai että vain yksi vastaanotin kerrallaan voi vastaanottaa. Edelleen tutkitaan WSRMax-ongelmaa laskevalla siirtotiellä OFDMA-järjestelmässä, ja johdetaan primaalihajotelmaan perustuva nopea algoritmi, joka yhteisoptimoi monen käyttäjän alikantoaalto- ja tehoallokaation maksimoiden painotetun summadatanopeuden. Numeeriset tulokset osoittavat, että esitetty algoritmi suppenee nopeammin kuin Lagrangen relaksaatioon perustuvat menetelmät. Lopuksi johdetaan hajautettu algoritmi WSRMax-ongelmalle monisoluisissa moniantennilähetystä käyttävissä järjestelmissä laskevaa siirtotietä varten. Esitetty menetelmä perustuu klassisiin primaalihajotelma- ja aligradienttimenetelmiin. Se ei turvaudu nollaanpakotus-keilanmuodostukseen tai korkean signaali-häiriö-plus-kohinasuhteen approksimaatioon, kuten monet muut hajautetut muunnelmat. Algoritmi koordinoi monta paikallista aliongelmaa (yhden kutakin tukiasemaa kohti) ratkaistakseen solujen välisen häiriön siten, että WSR maksimoituu. Numeeriset tulokset osoittavat, että merkittävää etua saadaan jo vähäisellä yhdessä toimivien tukiasemien välisellä viestinvaihdolla, vaikka globaalisti optimaalista ratkaisua ei voidakaan taata.
15

Algorithmes d'optimisation en grande dimension : applications à la résolution de problèmes inverses / Large scale optimization algorithms : applications to solution of inverse problems

Repetti, Audrey 29 June 2015 (has links)
Une approche efficace pour la résolution de problèmes inverses consiste à définir le signal (ou l'image) recherché(e) par minimisation d'un critère pénalisé. Ce dernier s'écrit souvent sous la forme d'une somme de fonctions composées avec des opérateurs linéaires. En pratique, ces fonctions peuvent n'être ni convexes ni différentiables. De plus, les problèmes auxquels on doit faire face sont souvent de grande dimension. L'objectif de cette thèse est de concevoir de nouvelles méthodes pour résoudre de tels problèmes de minimisation, tout en accordant une attention particulière aux coûts de calculs ainsi qu'aux résultats théoriques de convergence. Une première idée pour construire des algorithmes rapides d'optimisation est d'employer une stratégie de préconditionnement, la métrique sous-jacente étant adaptée à chaque itération. Nous appliquons cette technique à l'algorithme explicite-implicite et proposons une méthode, fondée sur le principe de majoration-minimisation, afin de choisir automatiquement les matrices de préconditionnement. L'analyse de la convergence de cet algorithme repose sur l'inégalité de Kurdyka-L ojasiewicz. Une seconde stratégie consiste à découper les données traitées en différents blocs de dimension réduite. Cette approche nous permet de contrôler à la fois le nombre d'opérations s'effectuant à chaque itération de l'algorithme, ainsi que les besoins en mémoire, lors de son implémentation. Nous proposons ainsi des méthodes alternées par bloc dans les contextes de l'optimisation non convexe et convexe. Dans le cadre non convexe, une version alternée par bloc de l'algorithme explicite-implicite préconditionné est proposée. Les blocs sont alors mis à jour suivant une règle déterministe acyclique. Lorsque des hypothèses supplémentaires de convexité peuvent être faites, nous obtenons divers algorithmes proximaux primaux-duaux alternés, permettant l'usage d'une règle aléatoire arbitraire de balayage des blocs. L'analyse théorique de ces algorithmes stochastiques d'optimisation convexe se base sur la théorie des opérateurs monotones. Un élément clé permettant de résoudre des problèmes d'optimisation de grande dimension réside dans la possibilité de mettre en oeuvre en parallèle certaines étapes de calculs. Cette parallélisation est possible pour les algorithmes proximaux primaux-duaux alternés par bloc que nous proposons: les variables primales, ainsi que celles duales, peuvent être mises à jour en parallèle, de manière tout à fait flexible. A partir de ces résultats, nous déduisons de nouvelles méthodes distribuées, où les calculs sont répartis sur différents agents communiquant entre eux suivant une topologie d'hypergraphe. Finalement, nos contributions méthodologiques sont validées sur différentes applications en traitement du signal et des images. Nous nous intéressons dans un premier temps à divers problèmes d'optimisation faisant intervenir des critères non convexes, en particulier en restauration d'images lorsque l'image originale est dégradée par un bruit gaussien dépendant du signal, en démélange spectral, en reconstruction de phase en tomographie, et en déconvolution aveugle pour la reconstruction de signaux sismiques parcimonieux. Puis, dans un second temps, nous abordons des problèmes convexes intervenant dans la reconstruction de maillages 3D et dans l'optimisation de requêtes pour la gestion de bases de données / An efficient approach for solving an inverse problem is to define the recovered signal/image as a minimizer of a penalized criterion which is often split in a sum of simpler functions composed with linear operators. In the situations of practical interest, these functions may be neither convex nor smooth. In addition, large scale optimization problems often have to be faced. This thesis is devoted to the design of new methods to solve such difficult minimization problems, while paying attention to computational issues and theoretical convergence properties. A first idea to build fast minimization algorithms is to make use of a preconditioning strategy by adapting, at each iteration, the underlying metric. We incorporate this technique in the forward-backward algorithm and provide an automatic method for choosing the preconditioning matrices, based on a majorization-minimization principle. The convergence proofs rely on the Kurdyka-L ojasiewicz inequality. A second strategy consists of splitting the involved data in different blocks of reduced dimension. This approach allows us to control the number of operations performed at each iteration of the algorithms, as well as the required memory. For this purpose, block alternating methods are developed in the context of both non-convex and convex optimization problems. In the non-convex case, a block alternating version of the preconditioned forward-backward algorithm is proposed, where the blocks are updated according to an acyclic deterministic rule. When additional convexity assumptions can be made, various alternating proximal primal-dual algorithms are obtained by using an arbitrary random sweeping rule. The theoretical analysis of these stochastic convex optimization algorithms is grounded on the theory of monotone operators. A key ingredient in the solution of high dimensional optimization problems lies in the possibility of performing some of the computation steps in a parallel manner. This parallelization is made possible in the proposed block alternating primal-dual methods where the primal variables, as well as the dual ones, can be updated in a quite flexible way. As an offspring of these results, new distributed algorithms are derived, where the computations are spread over a set of agents connected through a general hyper graph topology. Finally, our methodological contributions are validated on a number of applications in signal and image processing. First, we focus on optimization problems involving non-convex criteria, in particular image restoration when the original image is corrupted with a signal dependent Gaussian noise, spectral unmixing, phase reconstruction in tomography, and blind deconvolution in seismic sparse signal reconstruction. Then, we address convex minimization problems arising in the context of 3D mesh denoising and in query optimization for database management
16

Algoritmy pro vybrané geometrické problémy nad zonotopy a jejich aplikace v optimalizaci a v analýze dat / Algorithms for various geometric problems over zonotopes and their applications in optimization and data analysis

Rada, Miroslav January 2009 (has links)
The thesis unifies the most important author's results in the field of algorithms concerning zonotopes and their applications in optimization and statistics. The computational-geometric results consist of a new compact output-sensitive algorithm for enumerating vertices of a zonotope, which outperforms the rival algorithm with the same complexity-theoretic properties both theoretically and empirically, and a polynomial algorithm for arbitrarily precise approximation of a zonotope with the Löwner-John ellipsoid. In the application area, the thesis presents a result, which connects linear regression model with interval outputs with the zonotope matters. The usage of presented geometric algorithms for solving a nonconvex optimisation problem is also discussed.
17

NFDNA - um algoritmo para otimização não convexa e não diferenciável

Fernandes, Camila de Freitas 08 April 2016 (has links)
Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-06-16T17:52:10Z No. of bitstreams: 1 camiladefreitasfernandes.pdf: 740367 bytes, checksum: fac5ab7dcb039b31d587151b9a53fab1 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-07-13T14:25:13Z (GMT) No. of bitstreams: 1 camiladefreitasfernandes.pdf: 740367 bytes, checksum: fac5ab7dcb039b31d587151b9a53fab1 (MD5) / Made available in DSpace on 2016-07-13T14:25:13Z (GMT). No. of bitstreams: 1 camiladefreitasfernandes.pdf: 740367 bytes, checksum: fac5ab7dcb039b31d587151b9a53fab1 (MD5) Previous issue date: 2016-04-08 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Neste trabalho estudamos um algoritmo para solução de problemas de otimização irrestrita com funções não necessariamente convexas ou diferenciáveis, denominado Nonsmooth Feasible Direction Nonconvex Algorithm - NFDNA, e fazemos uma aplicação deste algoritmo que consistiu em utilizá-lo como subrotina de um outro algoritmo chamado Interior Epigraph Direction (IED) method. O IED, desenvolvido para resolver problemas de otimização não convexa, não diferenciável mas com restrições, utiliza Dualidade Lagrangeana que requer a minimização da função Lagrangeana. A eficiência do IED depende fortemente de tal minimização. Como aplicação, substituímos a rotina fminsearch do Matlab, utilizada originalmente pelo IED, pelo NFDNA. Mostramos através da solução de problemas teste que a performance do IED foi mais eficiente com a utilização do NFDNA. / In this work we study an algorithm for solving unsconstrained, not necessarily convex or differentiable optimization problems called Nonsmooth Feasible Direction Nonconvex Algorithm - NFDNA. We also employ this algorithm as a subroutine of the Interior Epigraph Directions (IED) method. The IED method, devised for solving constrained, nonconvex and nonsmooth optimization problems uses Lagrangean Duality which requires the minimization of the Lagrangean function. The effectiveness of the IED depends strongly on the Lagrangean function minimization. As an application, we replace the Matlab routine fminsearch, originally used by IED, with NFDNA. We show through the solution of test problems that the IED performance is more efficient by employing NFDNA.
18

Radio resource allocation techniques for MISO downlink cellular networks

Joshi, S. K. (Satya Krishna) 02 January 2018 (has links)
Abstract This thesis examines radio resource management techniques for multicell multi-input single-output (MISO) downlink networks. Specifically, the thesis focuses on developing linear transmit beamforming techniques by optimizing certain quality-of-service (QoS) features, including, spectral efficiency, fairness, and throughput. The problem of weighted sum-rate-maximization (WSRMax) has been identified as a central problem to many network optimization methods, and it is known to be NP-hard. An algorithm based on a branch and bound (BB) technique which globally solves the WSRMax problem with an optimality certificate is proposed. Novel bounding techniques via conic optimization are introduced and their efficiency is illustrated by numerical simulations. The proposed BB based algorithm is not limited to WSRMax only; it can be easily extended to maximize any system performance metric that can be expressed as a Lipschitz continuous and increasing function of the signal-to-interference-plus-noise (SINR) ratio. Beamforming techniques can provide higher spectral efficiency, only when the channel state information (CSI) of users is accurately known. However, in practice the CSI is not perfect. By using an ellipsoidal uncertainty model for CSI errors, both optimal and suboptimal robust beamforming techniques for the worst-case WSRMax problem are proposed. The optimal method is based on a BB technique. The suboptimal algorithm is derived using alternating optimization and sequential convex programming. Through a numerical example it is also shown how the proposed algorithms can be applied to a scenario with statistical channel errors. Next two decentralized algorithms for multicell MISO networks are proposed. The optimization problems considered are: P1) minimization of the total transmission power subject to minimum SINR constraints of each user, and P2) SINR balancing subject to the total transmit power constraint of the base stations. Problem P1 is of great interest for obtaining a transmission strategy with minimal transmission power that can guarantee QoS for users. In a system where the power constraint is a strict system restriction, problem P2 is useful in providing fairness among the users. Decentralized algorithms for both problems are derived by using a consensus based alternating direction method of multipliers. Finally, the problem of spectrum sharing between two wireless operators in a dynamic MISO network environment is investigated. The notion of a two-person bargaining problem is used to model the spectrum sharing problem, and it is cast as a stochastic optimization. For this problem, both centralized and distributed dynamic resource allocation algorithms are proposed. The proposed distributed algorithm is more suitable for sharing the spectrum between the operators, as it requires a lower signaling overhead, compared with centralized one. Numerical results show that the proposed distributed algorithm achieves almost the same performance as the centralized one. / Tiivistelmä Tässä väitöskirjassa tarkastellaan monisoluisten laskevan siirtotien moniantennilähetystä käyttävien verkkojen radioresurssien hallintatekniikoita. Väitöskirjassa keskitytään erityisesti kehittämään lineaarisia siirron keilanmuodostustekniikoita optimoimalla tiettyjä palvelun laadun ominaisuuksia, kuten spektritehokkuutta, tasapuolisuutta ja välityskykyä. Painotetun summadatanopeuden maksimoinnin (WSRMax) ongelma on tunnistettu keskeiseksi monissa verkon optimointitavoissa ja sen tiedetään olevan NP-kova. Tässä työssä esitetään yleinen branch and bound (BB) -tekniikkaan perustuva algoritmi, joka ratkaisee WSRMax-ongelman globaalisti ja tuottaa todistuksen ratkaisun optimaalisuudesta. Samalla esitellään uusia conic-optimointia hyödyntäviä suorituskykyrajojen laskentatekniikoita, joiden tehokkuutta havainnollistetaan numeerisilla simuloinneilla. Ehdotettu BB-perusteinen algoritmi ei rajoitu pelkästään WSRMax-ongelmaan, vaan se voidaan helposti laajentaa maksimoimaan mikä tahansa järjestelmän suorituskykyarvo, joka voidaan ilmaista Lipschitz-jatkuvana ja signaali-(häiriö+kohina) -suhteen (SINR) kasvavana funktiona. Keilanmuodostustekniikat voivat tuottaa suuremman spektritehokkuuden vain, jos käyttäjien kanavien tilatiedot tiedetään tarkasti. Käytännössä kanavan tilatieto ei kuitenkaan ole täydellinen. Tässä väitöskirjassa ehdotetaan WSRMax-ongelman ääritapauksiin sekä optimaalinen että alioptimaalinen keilanmuodostustekniikka soveltaen tilatietovirheisiin ellipsoidista epävarmuusmallia. Optimaalinen tapa perustuu BB-tekniikkaan. Alioptimaalinen algoritmi johdetaan peräkkäistä konveksiohjelmointia käyttäen. Numeerisen esimerkin avulla näytetään, miten ehdotettuja algoritmeja voidaan soveltaa skenaarioon, jossa on tilastollisia kanavavirheitä. Seuraavaksi ehdotetaan kahta hajautettua algoritmia monisoluisiin moniantennilähetyksellä toimiviin verkkoihin. Tarkastelun kohteena olevat optimointiongelmat ovat: P1) lähetyksen kokonaistehon minimointi käyttäjäkohtaisten minimi-SINR-rajoitteiden mukaan ja P2) SINR:n tasapainottaminen tukiasemien kokonaislähetystehorajoitusten mukaisesti. Ongelma P1 on erittäin kiinnostava, kun pyritään kehittämään mahdollisimman pienen lähetystehon vaativa lähetysstrategia, joka pystyy takaamaan käyttäjien palvelun laadun. Ongelma P2 on hyödyllinen tiukasti tehorajoitetussa järjestelmässä, koska se tarjoaa tasapuolisuutta käyttäjien välillä. Molempien ongelmien hajautetut algoritmit johdetaan konsensusperusteisen vuorottelevan kertoimien suuntaustavan avulla. Lopuksi tarkastellaan kahden langattoman operaattorin välisen spektrinjaon ongelmaa dynaamisessa moniantennilähetystä käyttävässä verkkoympäristössä. Spektrinjako-ongelmaa mallinnetaan käyttämällä kahden osapuolen välistä neuvottelua stokastisen optimoinnin näkökulmasta. Tähän ongelmaan ehdotetaan ratkaisuksi sekä keskitettyä että hajautettua resurssien allokoinnin algoritmia. Hajautettu algoritmi sopii paremmin spektrin jakamiseen operaattorien välillä, koska se vaatii vähemmän kontrollisignalointia. Numeeriset tulokset osoittavat, että ehdotetulla hajautetulla algoritmilla saavutetaan lähes sama suorituskyky kuin keskitetyllä algoritmillakin.
19

Modélisation et techniques d'optimisation en bio-informatique et fouille de données / Modelling and techniques of optimization in bioinformatics and data mining

Belghiti, Moulay Tayeb 01 February 2008 (has links)
Cette thèse est particulièrement destinée à traiter deux types de problèmes : clustering et l'alignement multiple de séquence. Notre objectif est de résoudre de manière satisfaisante ces problèmes globaux et de tester l'approche de la Programmation DC et DCA sur des jeux de données réelles. La thèse comporte trois parties : la première partie est consacrée aux nouvelles approches de l'optimisation non convexe. Nous y présentons une étude en profondeur de l'algorithme qui est utilisé dans cette thèse, à savoir la programmation DC et l'algorithme DC (DCA). Dans la deuxième partie, nous allons modéliser le problème clustering en trois sous-problèmes non convexes. Les deux premiers sous-problèmes se distinguent par rapport au choix de la norme utilisée, (clustering via les normes 1 et 2). Le troisième sous-problème utilise la méthode du noyau, (clustering via la méthode du noyau). La troisième partie sera consacrée à la bio-informatique. On va se focaliser sur la modélisation et la résolution de deux sous-problèmes : l'alignement multiple de séquence et l'alignement de séquence d'ARN par structure. Tous les chapitres excepté le premier se terminent par des tests numériques. / This Ph.D. thesis is particularly intended to treat two types of problems : clustering and the multiple alignment of sequence. Our objective is to solve efficiently these global problems and to test DC Programming approach and DCA on real datasets. The thesis is divided into three parts : the first part is devoted to the new approaches of nonconvex optimization-global optimization. We present it a study in depth of the algorithm which is used in this thesis, namely the programming DC and the algorithm DC ( DCA). In the second part, we will model the problem clustering in three nonconvex subproblems. The first two subproblems are distinguished compared to the choice from the norm used, (clustering via norm 1 and 2). The third subproblem uses the method of the kernel, (clustering via the method of the kernel). The third part will be devoted to bioinformatics, one goes this focused on the modeling and the resolution of two subproblems : the multiple alignment of sequence and the alignment of sequence of RNA. All the chapters except the first end in numerical tests.
20

Parallel and Decentralized Algorithms for Big-data Optimization over Networks

Amir Daneshmand (11153640) 22 July 2021 (has links)
<p>Recent decades have witnessed the rise of data deluge generated by heterogeneous sources, e.g., social networks, streaming, marketing services etc., which has naturally created a surge of interests in theory and applications of large-scale convex and non-convex optimization. For example, real-world instances of statistical learning problems such as deep learning, recommendation systems, etc. can generate sheer volumes of spatially/temporally diverse data (up to Petabytes of data in commercial applications) with millions of decision variables to be optimized. Such problems are often referred to as Big-data problems. Solving these problems by standard optimization methods demands intractable amount of centralized storage and computational resources which is infeasible and is the foremost purpose of parallel and decentralized algorithms developed in this thesis.</p><p><br></p><p>This thesis consists of two parts: (I) Distributed Nonconvex Optimization and (II) Distributed Convex Optimization.</p><p><br></p><p>In Part (I), we start by studying a winning paradigm in big-data optimization, Block Coordinate Descent (BCD) algorithm, which cease to be effective when problem dimensions grow overwhelmingly. In particular, we considered a general family of constrained non-convex composite large-scale problems defined on multicore computing machines equipped with shared memory. We design a hybrid deterministic/random parallel algorithm to efficiently solve such problems combining synergically Successive Convex Approximation (SCA) with greedy/random dimensionality reduction techniques. We provide theoretical and empirical results showing efficacy of the proposed scheme in face of huge-scale problems. The next step is to broaden the network setting to general mesh networks modeled as directed graphs, and propose a class of gradient-tracking based algorithms with global convergence guarantees to critical points of the problem. We further explore the geometry of the landscape of the non-convex problems to establish second-order guarantees and strengthen our convergence to local optimal solutions results to global optimal solutions for a wide range of Machine Learning problems.</p><p><br></p><p>In Part (II), we focus on a family of distributed convex optimization problems defined over meshed networks. Relevant state-of-the-art algorithms often consider limited problem settings with pessimistic communication complexities with respect to the complexity of their centralized variants, which raises an important question: can one achieve the rate of centralized first-order methods over networks, and moreover, can one improve upon their communication costs by using higher-order local solvers? To answer these questions, we proposed an algorithm that utilizes surrogate objective functions in local solvers (hence going beyond first-order realms, such as proximal-gradient) coupled with a perturbed (push-sum) consensus mechanism that aims to track locally the gradient of the central objective function. The algorithm is proved to match the convergence rate of its centralized counterparts, up to multiplying network factors. When considering in particular, Empirical Risk Minimization (ERM) problems with statistically homogeneous data across the agents, our algorithm employing high-order surrogates provably achieves faster rates than what is achievable by first-order methods. Such improvements are made without exchanging any Hessian matrices over the network. </p><p><br></p><p>Finally, we focus on the ill-conditioning issue impacting the efficiency of decentralized first-order methods over networks which rendered them impractical both in terms of computation and communication cost. A natural solution is to develop distributed second-order methods, but their requisite for Hessian information incurs substantial communication overheads on the network. To work around such exorbitant communication costs, we propose a “statistically informed” preconditioned cubic regularized Newton method which provably improves upon the rates of first-order methods. The proposed scheme does not require communication of Hessian information in the network, and yet, achieves the iteration complexity of centralized second-order methods up to the statistical precision. In addition, (second-order) approximate nature of the utilized surrogate functions, improves upon the per-iteration computational cost of our earlier proposed scheme in this setting.</p>

Page generated in 0.664 seconds