My thesis work started in the summer of 2005 as a three way joint project by Professor Castro and Mr. John Kwon and myself. A paper from this joint project was written and the content now forms my thesis.

Submitted by Marlene Santos (marlene.bc.ufg@gmail.com) on 2014-08-29T19:24:13Z No. of bitstreams: 2 Dissertação Corrigida e Finalisada.pdf: 2280692 bytes, checksum: fa3c7d92b5ed8a39139ceeb3abb80551 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2014-08-29T19:24:13Z (GMT). No. of bitstreams: 2 Dissertação Corrigida e Finalisada.pdf: 2280692 bytes, checksum: fa3c7d92b5ed8a39139ceeb3abb80551 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2013-12-13 / Conselho Nacional de Pesquisa e Desenvolvimento Científico e Tecnológico - CNPq / In this dissertation we analyze questions of existence of a weak solution for a class of superlineares elliptic Dirichlet problems. Here we do not consider the Ambrosseti Rabinovitz condition , which restricts some nonlinearities. We obtain main results of this dissertation via Variational Methods, such as Mountain Pass Theorem and Linking Theorem. Furthermore, weusePalais-Smalecondition(P.S.) or Cerami condition(Ce) / Nesta dissertação analisamos questões de existência de uma solução fraca para uma classe de problemas de Dirichlet elípticos superlineares. Aqui não consideramos a condição deAmbrosetti-Rabinowitz,a qual restringealgumasfunçõesnão lineares. Obtemos os principais resultados desta dissertação via Métodos variacionais, tais como o Teorema do Passo da Montanha e um Teorema de Linking. Além disso, utilizamos a TeoriaEspectral e ascondições dePalais-Smale(P.S.) eCerami(Ce).

In this paper we consider a radially symmetric nonlinear Dirichlet problem in a ball, where the nonlinearity is "superlinear" and "superlinear with jumping."

We establish the existence of radial solutions to the p-Laplacian equation ∆p u + f(u)=0 in RN, where f behaves like |u|q-1 u when u is large and f(u) < 0 for small positive u. We show that for each nonnegative integer n, there is a localized solution u which has exactly n zeros. Also, we look for radial solutions of a superlinear Dirichlet problem in a ball. We show that for each nonnegative integer n, there is a solution u which has exactly n zeros. Here we give an alternate proof to that which was given by Castro and Kurepa.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19122号 / 情博第568号 / 新制||情||100(附属図書館) / 32073 / 京都大学大学院情報学研究科数理工学専攻 / (主査)教授 山下 信雄, 教授 太田 快人, 教授 永持 仁 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

In dieser Arbeit präsentieren wir eine neue Art von Newton-Verfahren mit Liniensuche, basierend auf Interpolation im Bildbereich nach Wedin et al. [LW84]. Von dem resultierenden stabilisierten Newton-Algorithmus wird theoretisch und praktisch gezeigt, dass er effizient ist im Falle von nichtsingulären Lösungen. Darüber hinaus wird beobachtet, dass er eine superlineare Rate von Konvergenz bei einfachen Singularitäten erhält. Hingegen ist vom Newton-Verfahren ohne Liniensuche bekannt, dass es nur linear von fast allen Punkten in der Nähe einer singulären Lösung konvergiert. In Hinsicht auf Anwendungen auf Komplementaritätsprobleme betrachten wir auch Systeme, deren Jacobimatrix nicht differenzierbar sondern nur semismooth ist. Auch hier erreicht unser stabilisiertes und beschleunigtes Newton- Verfahren Superlinearität bei einfachen Singularitäten. / In this thesis we present a new type of line-search for Newton’s method, based on range space interpolation as suggested by Wedin et al. [LW84]. The resulting stabilized Newton algorithm is theoretically and practically shown to be efficient in the case of nonsingular roots. Moreover it is observed that it maintains a superlinear rate of convergence at simple singularities. Whereas Newton’s method without line-search is known to converge only linearly from almost all points near the singular root. In view of applications to complementarity problems we also consider systems, whose Jacobian is not differentiable but only semismooth. Again, our stabilized and accelerated Newton’s method achieves superlinearity at simple singularities.

Cette thèse est constituée de trois parties. Dans la première partie, on s’intéresse au problème de trace au bord d’une solution positive de l’équation (E1) - Δu + g(∇u) = 0 dans un domaine borné Ω. Si g(r) ≥ rq avec q > 1, on prouve que toute solution positive de (E1)admet une trace au bord considérée comme une mesure de Borel régulière. Si g(r) = rq avec1 < q < qc = N+1/N , on montre l’existence d’une solution positive dont la trace au bord est une mesure de Borel régulière. Si g(r) = rq avec qc ≤ q < 2, on établit une condition nécessaire de résolution en terme de capacité de Bessel C2-q/q ,q’ . On étudie aussi des ensembles éliminables au bord pour des solutions modérées et sigma-modérées. La deuxième partie est consacrée à étudier la limite, lorsque k → ∞, de solutions d’équation ∂tu - Δu + f(u) = 0 dans ℝN × (0,∞) avec donnée initiale kδ0. On prouve qu’il existe essentiellement trois types de comportement possible et démontre un résultat général d’existence de trace initiale et quelques résultats d’unicité et de non-unicité de solutions dont la donnée initiale n’est pas bornée. Dans la troisième partie, on considère l’équation ∂tu - Δu + f(u) = 0 dans ℝN × (0,∞) où p > 1. Si p > 2N/N+1, on fournit une condition suffisante portant sur f pour l’existence et l’unicité des solutions fondamentales et on étudie la limite lorsque k → ∞. On donne aussi de nouveaux résultats de non-unicité de solutions avec donnée initiale non bornée. Si p ≥ 2, on prouve que toute solution positive admet une trace initiale dans la classe des mesures de Borel régulières positives. Finalement on applique les résultats ci-dessus au cas f(u) = uα lnβ(u + 1) avec α,β > 0. / This thesis is divided into three parts. In the first part, we study the boundary trace of positive solutions of the equation (E1) - Δu + g(∇u) = 0 in a bounded domain . When g(r) ≥ rq with q > 1, we prove that any positive function of (E1) admits a boundary trace which is an outer regular Borel measure. When g(r) ≥ rq with 1 < q < qc = N+1/N, we prove the existence of a positive solution with a general outer regular Borel measure as boundary trace.When g(r) ≥ rq with qc ≤ q < 2, we establish a necessary condition for solvability in term of the Bessel capacity C2-q/q ,q’ . We also study boundary removable sets for moderate and sigma-moderate solutions. The second part is devoted to investigate the limit, when k → ∞, of the solutions of ∂tu - Δu + f(u) = 0 in ℝN × (0,∞) with initial data kδ0. We prove that there exist essentially three types of possible behaviour and provide a new and more general construction of the initial trace and some uniqueness and non-uniqueness results for solutions with unbounded initial data. In the third part, we consider the equation ∂tu - Δu + f(u) = 0 in ℝN × (0,∞) where p > 1. If p > 2N/N+1we provide a sufficient condition on f for existence and uniqueness of the fundamental solutions and we study their limit when k → ∞. We also give new results dealing with non uniqueness for the initial value problem with unbounded initial data. If p ≥ 2, we prove that any positive solution admits an initial trace in the class of positive Borel measures. Finally we apply the above results to the case f(u) = uα lnβ(u + 1) with α,β > 0.

Submitted by Johnny Rodrigues (johnnyrodrigues@ufcg.edu.br) on 2018-07-22T14:08:51Z No. of bitstreams: 1 FRANCISCA LEIDMAR JOSUÉ VIEIRA - DISSERTAÇÃO PPGMAT 2009..pdf: 290173 bytes, checksum: 21a058b9a6d5dfdd80b44bb2f900d25f (MD5) / Made available in DSpace on 2018-07-22T14:08:51Z (GMT). No. of bitstreams: 1 FRANCISCA LEIDMAR JOSUÉ VIEIRA - DISSERTAÇÃO PPGMAT 2009..pdf: 290173 bytes, checksum: 21a058b9a6d5dfdd80b44bb2f900d25f (MD5) Previous issue date: 2009-03 / Capes / O resumo foi escrito utilizando formulas e equações matemáticas que não fora possíveis serem transcritas aqui. Para a visualizar o resumo recomendamos o downloado do arquivo. / The abstract was written using mathematical formulas and equations that could not be transcribed here. To view the summary we recommend downloading the file.

We present a unified and self-contained treatment of rational Krylov methods for approximating the product of a function of a linear operator with a vector. With the help of general rational Krylov decompositions we reveal the connections between seemingly different approximation methods, such as the Rayleigh–Ritz or shift-and-invert method, and derive new methods, for example a restarted rational Krylov method and a related method based on rational interpolation in prescribed nodes. Various theorems known for polynomial Krylov spaces are generalized to the rational Krylov case. Computational issues, such as the computation of so-called matrix Rayleigh quotients or parallel variants of rational Arnoldi algorithms, are discussed. We also present novel estimates for the error arising from inexact linear system solves and the approximation error of the Rayleigh–Ritz method. Rational Krylov methods involve several parameters and we discuss their optimal choice by considering the underlying rational approximation problems. In particular, we present different classes of optimal parameters and collect formulas for the associated convergence rates. Often the parameters leading to best convergence rates are not optimal in terms of computation time required by the resulting rational Krylov method. We explain this observation and present new approaches for computing parameters that are preferable for computations. We give a heuristic explanation of superlinear convergence effects observed with the Rayleigh–Ritz method, utilizing a new theory of the convergence of rational Ritz values. All theoretical results are tested and illustrated by numerical examples. Numerous links to the historical and recent literature are included.

Dans cette thèse, nous nous étudions des algorithmes d’optimisation non linéaire. D’une part nous proposons des techniques de détection rapide de la non-réalisabilité d’un problème à résoudre. D’autre part, nous analysons le comportement local des algorithmes pour la résolution de problèmes singuliers. Dans la première partie, nous présentons une modification d’un algorithme de lagrangien augmenté pour l’optimisation avec contraintes d’égalité. La convergence quadratique du nouvel algorithme dans le cas non-réalisable est démontrée théoriquement et numériquement. La seconde partie est dédiée à l’extension du résultat précédent aux problèmes d’optimisation non linéaire généraux avec contraintes d’égalité et d’inégalité. Nous proposons une modification d’un algorithme de pénalisation mixte basé sur un lagrangien augmenté et une barrière logarithmique. Les résultats théoriques de l’analyse de convergence et quelques tests numériques montrent l’avantage du nouvel algorithme dans la détection de la non-réalisabilité. La troisième partie est consacrée à étudier le comportement local d’un algorithme primal-dual de points intérieurs pour l’optimisation sous contraintes de borne. L’analyse locale est effectuée sans l’hypothèse classique des conditions suffisantes d’optimalité de second ordre. Celle-ci est remplacée par une hypothèse plus faible basée sur la notion de borne d’erreur locale. Nous proposons une technique de régularisation de la jacobienne du système d’optimalité à résoudre. Nous démontrons ensuite des propriétés de bornitude de l’inverse de ces matrices régularisées, ce qui nous permet de montrer la convergence superlinéaire de l’algorithme. La dernière partie est consacrée à l’analyse de convergence locale de l’algorithme primal-dual qui est utilisé dans les deux premières parties de la thèse. En pratique, il a été observé que cet algorithme converge rapidement même dans le cas où les contraintes ne vérifient l’hypothèse de qualification de Mangasarian-Fromovitz. Nous démontrons la convergence superlinéaire et quadratique de cet algorithme, sans hypothèse de qualification des contraintes. / This thesis is devoted to the study of numerical algorithms for nonlinear optimization. On the one hand, we propose new strategies for the rapid infeasibility detection. On the other hand, we analyze the local behavior of primal-dual algorithms for the solution of singular problems. In the first part, we present a modification of an augmented Lagrangian algorithm for equality constrained optimization. The quadratic convergence of the new algorithm in the infeasible case is theoretically and numerically demonstrated. The second part is dedicated to extending the previous result to the solution of general nonlinear optimization problems with equality and inequality constraints. We propose a modification of a mixed logarithmic barrier-augmented Lagrangian algorithm. The theoretical convergence results and the numerical experiments show the advantage of the new algorithm for the infeasibility detection. In the third part, we study the local behavior of a primal-dual interior point algorithm for bound constrained optimization. The local analysis is done without the standard assumption of the second-order sufficient optimality conditions. These conditions are replaced by a weaker assumption based on a local error bound condition. We propose a regularization technique of the Jacobian matrix of the optimality system. We then demonstrate some boundedness properties of the inverse of these regularized matrices, which allow us to prove the superlinear convergence of our algorithm. The last part is devoted to the local convergence analysis of the primal-dual algorithm used in the first two parts of this thesis. In practice, it has been observed that this algorithm converges rapidly even in the case where the constraints do not satisfy the Mangasarian-Fromovitz constraint qualification. We demonstrate the superlinear and quadratic convergence of this algorithm without any assumption of constraint qualification.