Dans cette thèse, nous étudions la complexité paramétrée de problèmes combinatoires dans les graphes. Plus précisément, nous présentons une multitude d’algorithmes de programmation dynamique ainsi que des réductions montrant que certains de ces algorithmes sont optimaux. Nous nous intéressons principalement à la treewidth, un paramètre de graphes pouvant être vu comme une mesure de distance entre la structure d’un graphe et la structure topologique d’un arbre. Certains de nos algorithmes sont aussi paramétrés par la taille de la solution demandée et le degré maximum du graphe donné en entrée. Nous avons obtenu un certain nombre de résultats dont certains d’entre eux sont listés ci-dessous. Nous présentons un encadrement du nombre de graphes étiquetés de treewidth bornée. Nous étendons le domaine d’application de la théorie de la bidimensionalité par contraction au delà des graphes ne contenant pas de graphe apex en tant que mineur. Nous montrons aussi que la technique des structures de Catalan, outil améliorant l’efficacité des algorithmes résolvant des problèmes de connexité lorsque le graphe d’entrée est creux, ne peut être appliquée à la totalité des problèmes de connectivité, même si l’on ne considère, parmi les graphes creux, que les graphes planaires. Nous considérons le problème F-M-Deletion qui, étant donné une collection de graphes F, un graphe G et un entier k, demande s’il est possible de retirer au plus k sommets de G de telle sorte que le graphe restant ne contienne aucun graphe de F en tant que mineur. Nous considérons aussi la version topologique de ce problème, à savoir F-TM-Deletion. Ces deux problèmes généralisent des problèmes de modification de graphes bien connus tels que Vertex Cover, Feedback Vertex Set et Vertex Planarization. En fonction de la collection de graphes F, nous utilisons différentes techniques de programmation dynamique pour résoudre F-M-Deletion et F-TM-Deletion paramétrés par la treewidth. Nous utilisons des techniques standards, la structure des graphes frontières et l’approche basée sur le rang. En dernier lieu, nous appliquons ces techniques algorithmiques à deux problèmes issus du réseau de communications, à savoir une variation du problème classique de domination et un problème consistant à trouver un arbre couvrant possédant certaines propriétés, et un problème issu de la bioinformatique consistant à construire un arbre contenant en tant que mineur (topologique) un ensemble d’arbres donnés correspondant à des relations d’évolution entre ensembles d’espèces. / In this thesis, we study the Parameterized Complexity of combinatorial problems on graphs. More precisely, we present a multitude of dynamic programming algorithms together with reductions showing optimality for some of them. We mostly deal with the graph parameter of treewidth, which can be seen as a measure of how close a graph is to the topological structure of a tree. We also parameterize some of our algorithms by two other parameters, namely the size of a requested solution and the maximum degree of the input graph. We obtain a number of results, some of which are listed in the following. We estimate the number of labeled graphs of bounded treewidth. We extend the horizon of applicability of the theory of contraction Bidimensionality further than apex-minor free graphs, leading to a wider applicability of the design of subexponential dynamic programming algorithms. We show that the Catalan structure technique, that is a tool used to improve algorithm efficiency for connectivity problems where the input graph is restricted to be sparse, cannot be applied to all planar connectivity problems. We consider the F-M-Deletion problem that, given a set of graphs F, a graph G, and an integer k, asks if we can remove at most k vertices from G such that the remaining graph does not contain any graph of F as a minor. We also consider the topological version of this problem, namely F-TM-Deletion. Both problems generalize some well-known vertex deletion problems, namely Vertex Cover, Feedback Vertex Set, and Vertex Planarization. Depending on the set F, we use distinct dynamic programming techniques to solve F-M-Deletion and F-TM-Deletion when parameterized by treewidth. Namely, we use standard techniques, the rank based approach, and the framework of boundaried graphs. Finally, we apply these techniques to two problems originating from Networks, namely a variation of the classical dominating set problem and a problem that consists in finding a spanning tree with specific properties, and to a problem from Bioinformatics, namely that of construcing a tree that contains as a minor (or topological minor) a set of given trees corresponding to the evolutionary relationships between sets of species.

In the present thesis we provide compact extended formulations for a wide range of polytopes associated with the constraint satisfaction problem (CSP), monadic second order logic (MSO) on graphs, and extensions of MSO, when the given instances have bounded treewidth. We show that our extended formulations have additional useful properties, and we uncover connections between MSO and CSP. We conclude that a combination of the MSO logic, CSP and geometry provides an extensible framework for the design of compact extended formulations and parameterized algorithms for graphs of bounded treewidth. Putting our framework to use, we settle the parameterized complexity landscape for various extensions of MSO when parameterized by two important graph width parameters, namely treewidth and neighborhood diversity. We discover that the (non)linearity of the MSO extension determines the difference between fixedparameter tractability and intractability when parameterized by neighborhood diversity. Finally, we study shifted combinatorial optimization, a new nonlinear optimization framework generalizing standard combinatorial optimization, and provide initial findings from the perspective of parameterized complexity

No description available.

The crossing number of a graph 𝐺 is the minimum number of crossings in any drawing of 𝐺 in the plane. The rectilinear crossing number of 𝐺 is the minimum number of crossings in any straight-line drawing of 𝐺. The Fáry-Wagner theorem states that planar graphs have rectilinear crossing number zero. By Wagner’s theorem, that is equivalent to stating that every graph that excludes 𝐾₅ and 𝐾₃,₃ as minors has rectilinear crossing number 0. We are interested in discovering other proper minor-closed families of graphs which admit strong upper bounds on their rectilinear crossing numbers. Unfortunately, it is known that the crossing number of 𝐾₃,ₙ with 𝑛 ≥ 1, which excludes 𝐾₅ as a minor, is quadratic in 𝑛, more specifically Ω(𝑛²). Since every 𝑛-vertex graph in a proper minor closed family has O(𝑛) edges, the rectilinear crossing number of all such graphs is trivially O(𝑛²). In fact, it is not hard to argue that O(𝑛) bound on the crossing number is the best one can hope for general enough proper minor-closed families of graphs and that to achieve O(𝑛) bounds, one has to both exclude a minor and bound the maximum degree of the graphs in the family. In this thesis, we do that for bounded degree graphs that exclude a single-crossing graph as a minor. A single-crossing graph is a graph whose crossing number is at most one. The main result of this thesis states that every graph 𝐺 that does not contain a single-crossing graph as a minor has a rectilinear crossing number O(∆𝑛), where 𝐺 has 𝑛 vertices and maximum degree ∆. This dependence on 𝑛 and ∆ is best possible. Note that each planar graph is a single-crossing graph, as is the complete graph 𝐾₅ and the complete bipartite graph 𝐾₃,₃. Thus, the result applies to 𝐾₅-minor-free graphs, 𝐾₃,₃-minor free graphs, as well as to bounded treewidth graphs. In the case of bounded treewidth graphs, the result improves on the previous best known bound of O(∆² · 𝑛) by Wood and Telle [New York Journal of Mathematics, 2007]. In the case of 𝐾₃,₃-minor free graphs, our result generalizes the result of Dujmovic, Kawarabayashi, Mohar and Wood [SCG 2008].

In the present work we study the problem of reconstructing a graph from its closed neighbourhood list. We will explore this problem, formulated by V. Sós, from the point of view of the fixed parameter complexity. We study the graph reconstruction problem in a more general setting, when the reconstructed graph is required to belong to some special graph class. In the present work we prove that this general problem lies in the complexity class FPT, when parametrized by the treewidth and maximum degree of the reconstructed graph, or by the number of certain special induced subgraphs if the reconstructed graph is 2-degenerate. Also, we prove that the graph reconstruction problem lies in the complexity class XP when parametrized by the vertex cover number. Finally, we prove mutual independence of the results

CoordenaÃÃo de AperfeÃoamento de Pessoal de NÃvel Superior / The concept of treewidth was introduced by Robertson and Seymour. Treewidth may be defined as the size of the largest vertex set in a tree decomposition. Recent results show that several NP-Complete problems can be solved in polynomial time, or linear, when restricted to graphs with small treewidth. In our bibliographic research, we focus attention on the calculation of lower bounds for the treewidth and we described, in our dissertation, some of the principal results already available in the literature. We realize that linear-integer formulations for determining the treewidth are very limited in the literature and there are no studies available on the polyhedra associated with them. The Elimination Order Formulation (EOF) has been proposed by Koster and Bodlaender. It is based on orderly disposal of vertices and the relationship between the treewidth of a graph and its chordalizations. As a result of our study, we present a simplification of EOF formulation, we show that the polyhedron associated with this simplification is affine isomorphic to the EOF formulation. We determine the dimension of the polyhedron associated with the simplification, we briefly present a set of very simple facets and we introduce, analyse and demonstrate be a facet, some more complex inequalities. / O conceito de largura em Ãrvore (âtreewidthâ) foi introduzido por Robertson e Seymour. A largura em Ãrvore de um grafo G à o mÃnimo k tal que G pode ser decomposto em uma DecomposiÃÃo em Ãrvore (DEA) com cada subconjunto de vÃrtice com no mÃximo k+1 vÃrtices. Resultados recentes demonstram que vÃrios problemas NP-Completos podem ser resolvidos em tempo polinomial, ou ainda linear, quando restritos a grafos com largura em Ãrvore pequena. Em nossa pesquisa bibliogrÃfica, focamos a atenÃÃo no cÃlculo de limites inferiores para a largura em Ãrvore e descrevemos, em nossa dissertaÃÃo, alguns dos resultados jà disponÃveis na literatura. NÃs percebemos que formulaÃÃes lineares-inteiras para a determinaÃÃo da largura em Ãrvore sÃo limitadas na literatura e nÃo hà estudos disponÃveis sobre os poliedros associados a elas. A formulaÃÃo por ordem de eliminaÃÃo (EOF) foi proposta por Koster e Bodlaender. Ela à baseada na eliminaÃÃo ordenada de vÃrtices e na relaÃÃo entre a largura em Ãrvore de um grafo e suas cordalizaÃÃes. Como resultado de nosso estudo, apresentamos uma simplificaÃÃo da formulaÃÃo EOF, demonstramos que o poliedro associado a simplificaÃÃo à afim-isomÃrfico ao da formulaÃÃo EOF, verificamos a dimensÃo do poliedro associado à simplificaÃÃo, apresentamos brevemente um rol de facetas muito simples desse poliedro e, em seguinte, introduzimos, analisamos e demonstramos ser faceta algumas desigualdades mais complexas.

Este trabalho fornece uma avaliação empírica do desempenho de Redes Bayesianas quando se impõe restrições à largura de árvore de sua estrutura. O desempenho da rede é visto especificamente pela sua capacidade de generalização e também pela precisão da inferência em problemas de tomada de decisão. Resultados preliminares sugerem que adicionar essa restrição na largura de árvore diminui a capacidade de generalização do modelo além de tornar a tarefa de aprendizado mais difícil. / This work provides an empirical evaluation of the performance of Bayesian Networks when treewidth is bounded. The performance of the network is viewed as its generalizability and also as the accuracy of inference in decision making problems. Preliminary results suggest that adding constraints to treewidth decreases the model performance on unseen data and makes the corresponding optimization problem more difficult.

Este trabalho fornece uma avaliação empírica do desempenho de Redes Bayesianas quando se impõe restrições à largura de árvore de sua estrutura. O desempenho da rede é visto especificamente pela sua capacidade de generalização e também pela precisão da inferência em problemas de tomada de decisão. Resultados preliminares sugerem que adicionar essa restrição na largura de árvore diminui a capacidade de generalização do modelo além de tornar a tarefa de aprendizado mais difícil. / This work provides an empirical evaluation of the performance of Bayesian Networks when treewidth is bounded. The performance of the network is viewed as its generalizability and also as the accuracy of inference in decision making problems. Preliminary results suggest that adding constraints to treewidth decreases the model performance on unseen data and makes the corresponding optimization problem more difficult.

對於任何一個圖G = (V;E) ，如果我們可以找到最少的k 個弦圖(V;Ei)，使得E = E1 \ \ Ek ，則我們定義此圖G = (V;E) 的chordality為k ；而一個圖G = (V;E) 的樹寬則被定義為此圖所有的樹分解的寬的最小值。在這篇論文中，最主要的結論是所有圖的chordality 會小於或等於它的樹寬；更特別的是，有一些平面圖的chordality 為3，而所有系列平行圖的chordality 頂多為2。 / The chordality of a graph G = (V;E) is dened as the minimum k such that we can write E = E1 \ \ Ek, where each (V;Ei) is a chordal graph. The treewidth of a graph G = (V;E) is dened to be the minimum width over all tree decompositions of G. In this thesis, the principal result is that the chordality of a graph is at most its treewidth. In particular, there are planar graphs with chordality 3, and series-parallel graphs have chordality at most 2.

In this thesis we study the isoperimetric problem on trees and graphs with bounded treewidth. Let G = (V,E) be a finite, simple and undirected graph. For let δ(S,G)= {(u,v) ε E : u ε S and v ε V – S }be the edge boundary of S. Given an integer i, 1 ≤ i ≤ | V| , let the edge isoperimetric value of G at I be defined as be(i,G)= mins v;|s|= i | δ(S,G)|. For S V, let φ(S,G) = {u ε V – S : ,such that be the vertex boundary of S. Given an integer i, 1 ≤ i ≤ | V| , let the vertex isoperimetric value of G at I be defined as bv(i,G)= The edge isoperimetric peak of G is defined as be(G) =. Similarly the vertex isoperimetric peak of G is defined as bv(G)= .The problem of determining a lower bound for the vertex isoperimetric peak in complete k-ary trees of depth d,Tdkwas recently considered in[32]. In the first part of this thesis we provide lower bounds for the edge and vertex isoperimetric peaks in complete k-ary trees which improve those in[32]. Our results are then generalized to arbitrary (rooted)trees. Let i be an integer where . For each i define the connected edge isoperimetric value and the connected vertex isoperimetric value of G at i as follows: is connected and is connected A meta-Fibonacci sequence is given by the reccurence a(n)= a(x1(n)+ a1′(n-1))+ a(x2(n)+ a2′(n -2)), where xi: Z+ → Z+ , i =1,2, is a linear function of n and ai′(j)= a(j) or ai′(j)= -a(j), for i=1,2. Sequences belonging to this class have been well studied but in general their properties remain intriguing. In the second part of the thesis we show an interesting connection between the problem of determining and certain meta-Fibonacci sequences. In the third part of the thesis we study the problem of determining be and bv algorithmically for certain special classes of graphs. Definition 0.1. A tree decomposition of a graph G = (V,E) is a pair where I is an index set, is a collection of subsets of V and T is a tree whose node set is I such that the following conditions are satisfied: (For mathematical equations pl see the pdf file)