An area in graph theory is graph colouring, which essentially is a labeling of the vertices or edges according to certain constraints. In this thesis we consider star edge colouring, which is a variant of proper edge colouring where we additionally require the graph to have no two-coloured paths or cycles with length 4. The smallest number of colours needed to colour a graph G with a star edge colouring is called the star chromatic index of G and is denoted <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?%5Cchi'_%7Bst%7D(G)" />. This paper proves an upper bound of the star chromatic index of bipartite graphs in terms of the maximum degree; the maximum degree of G is the largest number of edges incident to a single vertex in G. For bipartite graphs Bk with maximum degree <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?k%5Cgeq1" />, the star chromatic index is proven to satisfy<img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?%20%5Cchi'_%7Bst%7D(B_k)%20%5Cleq%20k%5E2%20-%20k%20+%201" />. For bipartite graphs <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?B_%7Bk,n%7D" />, where all vertices in one part have degree n, and all vertices in the other part have degree k, it is proven that the star chromatic index satisfies <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?%5Cchi'_%7Bst%7D(Bk,n)%20%5Cleq%20k%5E2%20-2k%20+%20n%20+%201,%20k%20%5Cgeq%20n%20%3E%201" />. We also prove an upper bound for a special case of multipartite graphs, namely <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?K_%7Bn,1,1,%5Cdots,1%7D" /> with m parts of size one. The star chromatic index of such a graph satisfies<img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?%5Cchi'_%7Bst%7D(K_%7Bn,1,1,%5Cdots,1%7D)%20%5Cleq%2015%5Clceil%5Cfrac%7Bn%7D%7B8%7D%5Crceil%5Ccdot%5Clceil%5Cfrac%7Bm%7D%7B8%7D%5Crceil%20+%20%5Cfrac%7B1%7D%7B2%7Dm(m-1),%5C,m%20%5Cgeq%205" />. For complete multipartite graphs where m &lt; 5, we prove lower upper bounds than the one above.

A degree-constrained subgraph construction (DCS) problem aims to find an optimal spanning subgraph (w.r.t an objective function) subject to certain degree constraints on the vertices. DCS generalizes many combinatorial optimization problems such as Matchings and Edge Covers and has many practical and real-world applications. This thesis focuses on DCS problems where there are only upper and lower bounds on the degrees, known as b-matching and b-edge cover problems, respectively. We explore linear and submodular functions as the objective functions of the subgraph construction.<br><br>The contributions of this thesis involve both the design of new approximation algorithms for these DCS problems, and also their applications to real-world contexts.<br>We designed, developed, and implemented several approximation algorithms for DCS problems. Although some of these problems can be solved exactly in polynomial time, often these algorithms are expensive, tedious to implement, and have little to no concurrency. On the contrary, many of the approximation algorithms developed here run in nearly linear time, are simple to implement, and are concurrent. Using the local dominance framework, we developed the first parallel algorithm submodular b-matching. For weighted b-edge cover, we improved the classic Greedy algorithm using the lazy evaluation technique. We also propose and analyze several approximation algorithms using the primal-dual linear programming framework and reductions to matching. We evaluate the practical performance of these algorithms through extensive experimental results.<br><br>The second contribution of the thesis is to utilize the novel algorithms in real-world applications. We employ submodular b-matching to generate a balanced task assignment for processors to build Fock matrices in the NWChemEx quantum chemistry software. Our load-balanced assignment results in a four-fold speedup per iteration of the Fock matrix computation and scales to 14,000 cores of the Summit supercomputer at Oak Ridge National Laboratory. Using approximate b-edge cover, we propose the first shared-memory and distributed-memory parallel algorithms for the adaptive anonymity problem. Minimum weighted b-edge cover and maximum weight b-matching are shown to be applicable to constructing graphs from datasets for machine learning tasks. We provide a mathematical optimization framework connecting the graph construction problem to the DCS problem.

<p>Studying properties of random polynomials has marked a shift in algebraic geometry. Instead of worst-case analysis, which often leads to overly pessimistic perspectives, randomness helps perform average-case analysis, and thus obtain a more realistic view. Also, via Erdos' astonishing 'probabilistic method', one can potentially obtain deterministic results by introducing randomness into a question that apriori had nothing to do with randomness. </p> <p><br></p> <p>In this thesis, we study topological questions in real algebraic geometry, o-minimal geometry and random algebraic geometry, with motivation from incidence combinatorics. Specifically, we prove results along two different threads:</p> <p><br></p> <p>1. Topology of semi-algebraic and definable (over any o-minimal structure over R) sets, in both deterministic and random settings.</p><p>2. Topology of random hypersurface arrangements. In this case, we also prove a result that could be of independent interest in random graph theory.</p> <p><br></p> <p>Towards the first thread, motivated by applications in o-minimal incidence combinatorics, we prove bounds (both deterministic and random) on the topological complexity (as measured by the Betti numbers) of general definable hypersurfaces restricted to algebraic sets. Given any sequence of hypersurfaces, we show that there exists a definable hypersurface G, and a sequence of polynomials, such that each manifold in the sequence of hypersurfaces appears as a component of G restricted to the zero set of some polynomial in the sequence of polynomials. This shows that the topology of the intersection of a definable hypersurface and an algebraic set can be made <i>arbitrarily pathological</i>. On the other hand, we show that for random polynomials, the Betti numbers of the restriction of the zero set of a random polynomial to any definable set deviates from a Bezout-type bound with <i>bounded probability</i>.</p> <p><br></p> <p>Progress in o-minimal incidence combinatorics has lagged behind the developments in incidence combinatorics in the algebraic case due to the absence of an o-minimal version of the Guth-Katz <i>polynomial partitioning</i> theorem, and the first part of our work explains why this is so difficult. However, our average result shows that if we can prove that the measure of the set of polynomials which satisfy a certain property necessary for polynomial partitioning is suitably bounded from below, by the <i>probabilistic method</i>, we get an o-minimal polynomial partitioning theorem. This would be a tremendous breakthrough and would enable progress on multiple fronts in model theoretic combinatorics. </p> <p><br></p> <p>Along the second thread, we have studied the average Betti numbers of <i>random hypersurface arrangements</i>. Specifically, we study how the average Betti numbers of a finite arrangement of random hypersurfaces grows in terms of the degrees of the polynomials in the arrangement, as well as the number of polynomials. This is proved using a random Mayer-Vietoris spectral sequence argument. We supplement this result with a better bound on the average Betti numbers when one considers an <i>arrangement of quadrics</i>. This question turns out to be equivalent to studying the expected number of connected components of a certain <i>random graph model</i>, which has not been studied before, and thus could be of independent interest. While our motivation once again was incidence combinatorics, we obtained the first bounds on the topology of arrangements of random hypersurfaces, with an unexpected bonus of a result in random graphs.</p>

If <i>E</i> is an exterior algebra on a finite dimensional vector space and <i>M</i> is a graded <i>E</i>-module, the relationship between the resonance varieties of <i>M</i> and the minimal free resolution of <i>M </i>is studied. In the context of Orlik–Solomon algebras, we give a condition under which elements of the second resonance variety can be obtained. We show that the resonance varieties of a general <i>M</i> are invariant under taking syzygy modules up to a shift. As corollary, it is shown that certain points in the resonance varieties of <i>M</i> can be determined from minimal syzygies of a special form, and we define syzygetic resonance varieties to be the subvarieties consisting of such points. The (depth one) syzygetic resonance varieties of a square-free module <i>M</i> over <i>E</i> are shown to be subspace arrangements whose components correspond to graded shifts in the minimal free resolution of <i>_SM</i>, the square-free module over a commutative polynomial ring <i>S </i>corresponding to <i>M</i>. Using this, a lower bound for the graded Betti numbers of the square-free module<i> M</i> is given. As another application, it is shown that the minimality of certain syzygies of Orlik–Solomon algebras yield linear subspaces of their (syzygetic) resonance varieties and lower bounds for their graded Betti numbers.

Based on the generating matrix to the Tribonacci sequence, the Tribonacci cat map is a discrete chaotic dynamical system, similar to Arnold's discrete cat map, but on three dimensional space. In this thesis, this new mapping is introduced and the properties of its matrix are presented. The main results of the investigation prove how the size of the domain of the map affects its period and explore the orbit lengths of non-trivial points. Different upper bounds to the map are studied and proved, and a conjecture based on numerical calculations is proposed. The Tribonacci cat map is used for applications such as 3D image encryption and colour encryption. In the latter case, the results provided by the mapping are compared to those from a generalised form of the map.

In this research,we study iterations of non-pleasantly ramified polynomials over fields of positive characteristic and subsequently, their lower ramification numbers. Of particular interest for this thesis are polynomials for which both the multiplicity and  the degree of its iterates grow exponentially. Specifically we consider the family  of polynomials such that given a positive integer k the family is given by P(z) = z(1 + z (3^k-1)/2 + z3^k-1). The cubic polynomial z + z2 + z3 is a special case of this family and is particularly interesting.

Many countries are reporting of difficulties in exact education at schools: mathematics, informatics, physics etc. Various methods are proposed to awaken and preserve students’ interest in these disciplines. Among them, the simplification, accent on applications, avoiding of argumentation (especially in mathematics) etc. must be mentioned. As one of reasons for these approaches the growing amount of knowledge/skills to be acquired at school is often mentioned. In this paper we consider one of the possibilities to integrate partially teaching of important chapters of discrete mathematics and informatics not reducing the high educational standards. The approach is based on the identification and mastering general combinatorial principles underlying many topics in both disciplines. A special attention in the paper is given to the so-called “pigeonhole principle” and its generalizations. In folklore, this principle is usually formulated in the following way: “if there are n + 1 rabbits in n cages, you can find a cage with at least two rabbits in it“. Examples of appearances of this principle both in mathematics and in computer science are considered.

In this workshop, we present a model of problem that we call Research Situation for the Classroom (RSC). The aim of a RSC is to put students in a role close to a mathematical researcher in order to make them work on mathematical thinking/skills. A RSC has some characteristics : the problem is close to a research one, the statement is an easy understandable question, school knowledge are elementary, there is no end, a solved question postponed to new questions... The most important characteristic of a RSC is that students can manage their research by fixing themselves some variable of the problem. So, a RSC is completely different from a problem that students usually do in France. For short : there is no final answer, students can try to resolve their own questions : a RSC is a large open field where many sub-problems exist; the goal for the students is not to apply a technique: the goal is, as for a researcher, to search. These type of situations are particularly interesting to develop problem solving skills and mathematical thinking. They can also let students discover that mathematics are “alive” and “realistic”. This workshop will be split into two parts. First, we propose to put people in the situation of solving a RSC to make them discover practically what is it. After, we present the model of a RSC and some results of our experimentations.

The orientation completion problem for a hereditary class C of oriented graphs asks whether a given partially oriented graph can be completed to a graph belonging to C. This problem was introduced recently and is a generalization of several existing problems, including the recognition problem for certain classes of graphs and the representation extension problem for proper interval graphs. A local tournament is an oriented graph in which the in-neighbourhood as well as the out-neighbourhood of each vertex induces a tournament. Local tournaments are a well-studied class of oriented graphs that generalize tournaments and their underlying graphs are intimately related to proper circular-arc graphs. Proper interval graphs are precisely those which can be oriented as acyclic local tournaments. The orientation completion problems for the class of local tournaments and the class of acyclic local tournaments have been shown to be polynomial-time solvable. In this thesis, we characterize the partially oriented graphs that can be completed to local tournaments by finding a complete list of obstructions. These are in a sense the minimal partially oriented graphs that cannot be completed to local tournaments. We also determine the minimal partially oriented graphs that cannot be completed to acyclic local tournaments. / Graduate

In this dissertation, we investigate various aspects of signed graphs, with a particular focus on signings and sign-symmetric signed graphs. We begin by examining the complete graph on six vertices with one edge deleted ($K_6$\textbackslash e) and explore the different ways of signing this graph up to switching isomorphism. We determine the frustration index (number) of these signings and investigate the existence of sign-symmetric signed graphs. We then extend our study to the $K_6$\textbackslash 2e graph and the McGee graph with exactly two negative edges. We investigate the distinct ways of signing these graphs up to switching isomorphism and demonstrate the absence of sign-symmetric signed graphs in some cases. We then introduce and study the signed graph class $\mathcal{S}$, which includes all sign-symmetric signed graphs, we prove several theorems and lemmas as well as discuss the class of tangled sign-symmetric signed graphs. Also, we study the graph class $\mathcal{G}$, consisting of graphs with at least one sign-symmetric signed graph, prove additional theorems and lemmas, and determine certain families within $\mathcal{G}$. Our results have practical applications in various fields such as social psychology and computer science.