Le spectre du Laplacien sans signe a fait l'objet de beaucoup d'attention dans la communauté scientifique ces dernières années. La principale raison est l'intuition, basée sur une étude des petits graphes et sur des propriétés valides pour des graphes de toutes tailles, que plus de graphes sont déterminés par le spectre de cette matrice que par celui de la matrice d'adjacence et du Laplacien. Les travaux présentés dans cette thèse ont apporté des éléments nouveaux sur les informations contenues dans le spectre cette matrice. D'une part, on y présente des relations entre les invariants de structure et une valeur propre du Laplacien sans signe. D'autre part, on présente des familles de graphes extrêmes pour deux de ses valeurs propres, avec et sans contraintes additionnelles sur la forme de graphe. Il se trouve que ceux-ci sont très similaires à ceux obtenus dans les mêmes conditions avec les valeurs propres de la matrice d'adjacence. Cela aboutit à la définition de familles de graphes pour lesquelles, le spectre du Laplacien sans signe ou une de ses valeurs propres, le nombre de sommets et un invariant de structure suffisent à déterminer le graphe. Ces résultats, par leur similitude avec ceux de la littérature viennent confirmer l'idée que le Laplacien sans signe détermine probablement aussi bien les graphes que la matrice d'adjacence.

Stable matchings were introduced in 1962 by David Gale and Lloyd Shapley to study the college admissions problem. The seminal work of Gale and Shapley has motivated hundreds of research papers and found applications in many areas of mathematics, computer science, economics, and even medicine. This thesis studies stable matchings in graphs and hypergraphs. We begin by introducing the work of Gale and Shapley. Their main contribution was the proof that every bipartite graph has a stable matching. Our discussion revolves around the Gale-Shapley algorithm and highlights some of the interesting properties of stable matchings in bipartite graphs. We then progress to non-bipartite graphs. Contrary to bipartite graphs, we may not be able to find a stable matching in a non-bipartite graph. Some of the work of Irving will be surveyed, including his extension of the Gale-Shapley algorithm. Irving's algorithm shows that many of the properties of bipartite stable matchings remain when the general case is examined. In 1991, Tan showed how to extend the fundamental theorem of Gale and Shapley to non-bipartite graphs. He proved that every graph contains a set of edges that is very similar to a stable matching. In the process, he found a characterization of graphs with stable matchings based on a modification of Irving's algorithm. Aharoni and Fleiner gave a non-constructive proof of Tan's Theorem in 2003. Their proof relies on a powerful topological result, due to Scarf in 1965. In fact, their result extends beyond graphs and shows that every hypergraph has a fractional stable matching. We show how their work provides new and simpler proofs to several of Tan's results. We then consider fractional stable matchings from a linear programming perspective. Vande Vate obtained the first formulation for complete bipartite graphs in 1989. Further, he showed that the extreme points of the solution set exactly correspond to stable matchings. Roth, Rothblum, and Vande Vate extended Vande Vate's work to arbitrary bipartite graphs. Abeledo and Rothblum further noticed that this new formulation can model fractional stable matchings in non-bipartite graphs in 1994. Remarkably, these formulations yield analogous results to those obtained from Gale-Shapley's and Irving's algorithms. Without the presence of an algorithm, the properties are obtained through clever applications of duality and complementary slackness. We will also discuss stable matchings in hypergraphs. However, the desirable properties that are present in graphs no longer hold. To rectify this problem, we introduce a new ``majority" stable matchings for 3-uniform hypergraphs and show that, under this stronger definition, many properties extend beyond graphs. Once again, the linear programming tools of duality and complementary slackness are invaluable to our analysis. We will conclude with a discussion of two open problems relating to stable matchings in 3-uniform hypergraphs.

Erdos proved the well-known result that every graph has a spanning, bipartite subgraph such that every vertex has degree at least half of its original degree. Bollobas and Scott conjectured that one can get a slightly weaker result if we require the subgraph to be not only spanning and bipartite, but also balanced. We prove this conjecture for graphs of maximum degree 3. The majority of the paper however, will focus on graph tiling. Graph tiling (or sometimes referred to as graph packing) is where, given a graph H, we find a spanning subgraph of some larger graph G that consists entirely of disjoint copies of H. With the Regularity Lemma and the Blow-up Lemma as our main tools, we prove an asymptotic minimum degree condition for an arbitrary bipartite graph G to be tiled by another arbitrary bipartite graph H. This proves a conjecture of Zhao and also implies an asymptotic version of a result of Kuhn and Osthus for bipartite graphs.

<p>Tree-like structures are fundamental in nature. A wide variety of two-dimensional imaging techniques allow us to image trees. However, an image of a tree typically includes spurious branch crossings and the original relationships of ancestry among edges may be lost. We present a methodology for estimating the most likely topology of a rooted, directed, three-dimensional tree given a single two-dimensional image of it. We regularize this inverse problem via a prior parametric tree-growth model that realistically captures the morphology of a wide variety of trees. We show that the problem of estimating the optimal tree has linear complexity if ancestry is known, but is NP-hard if it is lost. For the latter case, we present both a greedy approximation algorithm and a heuristic search algorithm that effectively explore the space of possible trees. Experimental results on retinal vessel, plant root, and synthetic tree datasets show that our methodology is both accurate and efficient.</p> / Dissertation

The GLuskap system for interactive three-dimensional graph drawing applies techniques of scientific visualization and interactive systems to the construction, display, and analysis of graph drawings. Important features of the system include support for large-screen stereographic 3D display with immersive head-tracking and motion-tracked interactive 3D wand control. A distributed rendering architecture contributes to the portability of the system, with user control performed on a laptop computer without specialized graphics hardware. An interface for implementing graph drawing layout and analysis algorithms in the Python programming language is also provided. This thesis describes comprehensively the work on the system by the author—this work includes the design and implementation of the major features described above. Further directions for continued development and research in cognitive tools for graph drawing research are also suggested. / viii, 110 leaves : ill. (some col.) ; 29 cm.

A 3D arrangement graph G is the abstract graph induced by an arrangement of planes in general position where the intersection of any two planes forms a line of intersection and an intersection of three planes creates a point. The properties of three classes of arrangement graphs — four, five and six planes — are investigated. For graphs induced from six planes, specialized methods were developed to ensure all possible graphs were discovered. The main results are: the number of 3D arrangement graphs induced by four, five and six planes are one, one and 43 respectively; the three classes are Hamiltonian; and the 3D arrangement graphs created from four and five planes are planar but none of the graphs created from six planes are planar. / x, 89 leaves : ill. (some col.) ; 29 cm

Solving NP-hard facility location problems in wireless network planning is a common scenario. In our research, we study the Covering problem, a well known facility location problem with applications in wireless network deployment. We focus on networks with a sparse structure. First, we analyzed two heuristics of building Tree Decomposition based on vertex separator and perfect elimination order. We extended the vertex separator heuristic to improve its time performance. Second, we propose a dynamic programming algorithm based on the Tree Decomposition to solve the Covering problem optimally on the network. We developed several heuristic techniques to speed up the algorithm. Experiment results show that one variant of the dynamic programming algorithm surpasses the performance of the state of the art mathematical optimization commercial software on several occasions. / ix, 89 leaves : ill. ; 29 cm

An independent set in a graph G is a set of pairwise nonadjacent vertices, and the maximum size, alpha(G), of an independent set in G is called the independence number. Given a graph G and weight matrix A of G with entries from some field F, the maximum dimension of an A-isotropic subspace, known as the Witt index of A, is an upper bound on alpha(G). Since any weight matrix can be used, it is natural to seek the minimum upper bound on the independence number of G that can be achieved by a weight matrix. This minimum, iota_F^*(G), is called the minimum Witt index of G over F, and the resulting bound, alpha(G)<= iota_F^*(G), is called the isotropic bound. When F is finite, the possible values of iota_F^*(G) are determined and the graphs that attain the isotropic bound are characterized. The characterization is given in terms of graph classes CC(n,t,c) and CK(n,t,k) constructed from certain spanning subgraphs called C(n,t,c)-graphs and K(n,t,k)-graphs. Here t is the term rank of the adjacency matrix of G. When F=R, the isotropic bound is known as the Cvetkovi\'c bound. It is shown that it is sufficient to consider a finite number of weight matrices A when determining iota_R^*(G) and that, in many cases, two weight values suffice. For example, if the vertex set of G can be covered by alpha(G) cliques, then G attains the Cvetkovi\'c bound with a weight matrix with two weight values. Inequalities on alpha and iota_F^* resulting from graph operations such as sums, products, vertex deletion, and vertex identification are examined and, in some cases, conditions that imply equality are proved. The equalities imply that the problem of determining whether or not alpha(G)=iota_F^*(G) can be reduced to that of determining iota_F^*(H) for certain crucial graphs H found from G. / Thesis (Ph.D, Mathematics & Statistics) -- Queen's University, 2007-09-04 15:38:47.57

In New Zealand, a formal tree improvement and breeding programme for Pinus radiata (D.Don) commenced in 1952. A countrywide series of progeny trials was progressively established on over seventy sites, and is managed by the Radiata Pine Breeding Company (RPBC). Diameter at breast height data from the series were used to investigate genotype x environment interaction with a view to establishing the need for partitioning breeding and deployment efforts for P. radiata. Nearly 300,000 measurements made this study one of the largest for genotype x environment interaction ever done. Bivariate analyses were conducted between all pairs of sites to determine genetic correlations between sites. Genetic correlations were used to construct a proximity matrix by subtracting each correlation from unity. The process of constructing the matrix highlighted issues of low connectivity between sites; whereby meaningful correlations between sites were established with just 5 % of the pairs. However, nearly two-thirds of these genetic correlations were between -1.0 and 0.6, indicating the presence of strong genotype x environment interactions. A technique known as multiple regression on resemblance matrices was carried out by regressing a number of environmental correlation matrices on the diameter at breast height correlation matrix. Genotype x environment interactions were found to be driven by extreme maximum temperatures (t-statistic of 2.03 against critical t-value of 1.96 at 95 % confidence level). When tested on its own, altitude was significant with genetic correlations between sites at the 90 % confidence level (t-statistic of 1.92 against critical t-value of 1.645). In addition, a method from Graph Theory using proximity thresholds was utilised as a form of clustering. However, this study highlighted the existence of high internal cohesion within trial series, and high external isolation between trial series. That is, grouping of sites (in terms of diameter) was observed to be a reflection of the series of trials for which each site was established. This characteristic is particularly unhelpful for partitioning sites into regions of similar propensity to genotype x environment interaction, as the genotype x environment effect is effectively over-ridden by the genotype effect. Better cohesion between past, present and future trial series, and more accurate bioclimatic data should allow more useful groupings of sites to be extracted from the data. Given this, however, it is clear that there are a large number of interactive families contained in the RPBC dataset. It is concluded that partitioning of New Zealand’s P. radiata breeding programme cannot be ruled out as an advantageous option.

Recently, many methods have been proposed for the classification and prediction problems in bioinformatics. One of these problems is the protein structure prediction. Machine learning approaches and new algorithms have been proposed to solve this problem. Among the machine learning approaches, Support Vector Machines (SVM) have attracted a lot of attention due to their high prediction accuracy. Since protein data consists of sequence and structural information, another most widely used approach for modeling this structured data is to use graphs. In computer science, graph theory has been widely studied; however it has only been recently applied to bioinformatics. In this work, we introduced new algorithms based on statistical methods, graph theory concepts and machine learning for the protein structure prediction problem. A new statistical method based on z-scores has been introduced for seed selection in proteins. A new method based on finding common cliques in protein data for feature selection is also introduced, which reduces noise in the data. We also introduced new binary classifiers for the prediction of structural transitions in proteins. These new binary classifiers achieve much higher accuracy results than the current traditional binary classifiers.