Yes / The work presented in this paper is motivated by a complex multivariate engineering problem associated with engine mapping experiments, which require efficient Design of Experiment (DoE) strategies to minimise expensive testing. The paper describes the development and evaluation of a Permutation Genetic Algorithm (PermGA) to support an exploration-based sequential DoE strategy for complex real-life engineering problems. A known PermGA was implemented to generate uniform OLH DoEs, and substantially extended to support generation of Model Building–Model Validation (MB-MV) sequences, by generating optimal infill sets of test points as OLH DoEs, that preserve good space filling and projection properties for the merged MB + MV test plan. The algorithm was further extended to address issues with non-orthogonal design spaces, which is a common problem in engineering applications. The effectiveness of the PermGA algorithm for the MB-MV OLH DoE sequence was evaluated through a theoretical benchmark problem based on the Six-Hump-Camel-Back (SHCB) function, as well as the Gasoline Direct Injection (GDI) engine steady state engine mapping problem that motivated this research. The case studies show that the algorithm is effective at delivering quasi-orthogonal space-filling DoEs with good properties even after several MB-MV iterations, while the improvement in model adequacy and accuracy can be monitored by the engineering analyst. The practical importance of this work, demonstrated through the engine case study, also is that significant reduction in the effort and cost of testing can be achieved. / The research work presented in this paper was funded by the UK Technology Strategy Board (TSB) through the Carbon Reduction through Engine Optimization (CREO) project.

Le système immunitaire adaptatif est capable de produire une réponse spécifique contre presque tous le pathogènes qui agressent notre organisme. Ceci est dû aux anticorps qui sont des protéines secrétées par les cellules B. Les molécules qui provoquent cette réaction sont appelées antigènes : pendant une réponse immunitaire, les cellules B sont soumises à un processus d’apprentissage afin d’améliorer leur capacité à reconnaitre un antigène donne. Ce processus est appelé maturation d’affinité des anticorps. Nous établissons un cadre mathématique très flexible dans lequel nous définissons et étudions des modelés évolutionnaires simplifies inspirés par la maturation d’affinité des anticorps. Nous identifions les éléments constitutifs fondamentaux de ce mécanisme d’évolution extrêmement rapide et efficace : mutation, division et sélection. En commençant par une analyse rigoureuse du mécanisme de mutation dans le Chapitre 2, nous procédons à l’enrichissement progressif du modelé en ajoutant et analysant le processus de division dans le Chapitre 3 ,puis des pressions sélectives dépendantes de l’affinité dans le Chapitre 4. Notre objectif n’est pas de construire un modèle mathématique très détaillé et exhaustif de la maturation d’affinité des anticorps, mais plutôt d’enquêter sur les interactions entre mutation, division et sélection dans un contexte théorique simplifie. On cherche à comprendre comment les différents paramètres biologiques influencent la fonctionnalité du système, ainsi qu’à estimer les temps caractéristiques de l’exploration de l’espace d’états des traits des cellules B. Au-delà des motivations biologiques de la modélisation de la maturation d’affinité des anticorps, l’analyse de ce processus d’apprentissage nous a amenée à concevoir un modèle mathématique qui peut également s’appliquer à d’autres systèmes d’évolution, mais aussi à l’étude de la propagation de rumeurs ou de virus. Notre travail théorique s’accompagne de nombreuses simulations numériques qui viennent soit l’illustrer soit montrer que certains résultats demeurent extensibles a des situations plus compliquées. / The adaptive immune system is able to produce a specific response against almost any pathogen that could penetrate our organism and inflict diseases. This task is assured by the production of antigen-specific antibodies secreted by B-cells. The agents which causes this reaction are called antigens: during an immune response B-cells are submitted to a learning process in order to improve their ability to recognize the immunizing antigen. This process is called antibody affinity maturation. We set a highly flexible mathematical environment in which we define and study simplified mathematical evolutionary models inspired by antibody affinity maturation. We identify the fundamental building blocks of this extremely efficient and rapid evolutionary mechanism: mutation, division and selection. Starting by a rigorous analysis of the mutational mechanism in Chapter 2, we proceed by successively enriching the model by adding and analyzing the division process in Chapter 3 and affinity-dependent selection pressures in Chapter 4. Our aim is not to build a very detailed and comprehensive mathematical model of antibody affinity maturation, but rather to investigate interactions between mutation, division and selection in a simplified theoretical context. We want to understand how the different biological parameters affect the system’s functionality, as well as estimate the typical time-scales of the exploration of the state-space of B-cell traits. Beyond the biological motivations of antibody affinity maturation modeling, the analysis of this learning process leads us to build a mathematical model which could be relevant to model other evolutionary systems, but also gossip or virus propagation. Our method is based on the complementarity between probabilistic tools and numerical simulations.

La simulation numérique est devenue une alternative à l’expérimentation réelle pour étudier des phénomènes physiques. Cependant, les phénomènes complexes requièrent en général un nombre important de simulations, chaque simulation étant très coûteuse en temps de calcul. Une approche basée sur la théorie des plans d’expériences est souvent utilisée en vue de réduire ce coût de calcul. Elle consiste à partir d’un nombre réduit de simulations, organisées selon un plan d’expériences numériques, à construire un modèle d’approximation souvent appelé métamodèle, alors beaucoup plus rapide à évaluer que le code lui-même. Traditionnellement, les plans utilisés sont des plans de type Space-Filling Design (SFD). La première partie de la thèse concerne la construction de plans d’expériences SFD à taille fixe adaptés à l’identification d’un modèle de krigeage car le krigeage est un des métamodèles les plus populaires. Nous étudions l’impact de la contrainte Hypercube Latin (qui est le type de plans les plus utilisés en pratique avec le modèle de krigeage) sur des plans maximin-optimaux. Nous montrons que cette contrainte largement utilisée en pratique est bénéfique quand le nombre de points est peu élevé car elle atténue les défauts de la configuration maximin-optimal (majorité des points du plan aux bords du domaine). Un critère d’uniformité appelé discrépance radiale est proposé dans le but d’étudier l’uniformité des points selon leur position par rapport aux bords du domaine. Ensuite, nous introduisons un proxy pour le plan minimax-optimal qui est le plan le plus proche du plan IMSE (plan adapté à la prédiction par krigeage) et qui est coûteux en temps de calcul, ce proxy est basé sur les plans maximin-optimaux. Enfin, nous présentons une procédure bien réglée de l’optimisation par recuit simulé pour trouver les plans maximin-optimaux. Il s’agit ici de réduire au plus la probabilité de tomber dans un optimum local. La deuxième partie de la thèse porte sur un problème légèrement différent. Si un plan est construit de sorte à être SFD pour N points, il n’y a aucune garantie qu’un sous-plan à n points (n 6 N) soit SFD. Or en pratique le plan peut être arrêté avant sa réalisation complète. La deuxième partie est donc dédiée au développement de méthodes de planification séquentielle pour bâtir un ensemble d’expériences de type SFD pour tout n compris entre 1 et N qui soient toutes adaptées à la prédiction par krigeage. Nous proposons une méthode pour générer des plans séquentiellement ou encore emboités (l’un est inclus dans l’autre) basée sur des critères d’information, notamment le critère d’Information Mutuelle qui mesure la réduction de l’incertitude de la prédiction en tout point du domaine entre avant et après l’observation de la réponse aux points du plan. Cette approche assure la qualité des plans obtenus pour toutes les valeurs de n, 1 6 n 6 N. La difficulté est le calcul du critère et notamment la génération de plans en grande dimension. Pour pallier ce problème une solution a été présentée. Cette solution propose une implémentation astucieuse de la méthode basée sur le découpage par blocs des matrices de covariances ce qui la rend numériquement efficace. / In recent years, computer simulation models are increasingly used to study complex phenomena. Such problems usually rely on very large sophisticated simulation codes that are very expensive in computing time. The exploitation of these codes becomes a problem, especially when the objective requires a significant number of evaluations of the code. In practice, the code is replaced by global approximation models, often called metamodels, most commonly a Gaussian Process (kriging) adjusted to a design of experiments, i.e. on observations of the model output obtained on a small number of simulations. Space-Filling-Designs which have the design points evenly spread over the entire feasible input region, are the most used designs. This thesis consists of two parts. The main focus of both parts is on construction of designs of experiments that are adapted to kriging, which is one of the most popular metamodels. Part I considers the construction of space-fillingdesigns of fixed size which are adapted to kriging prediction. This part was started by studying the effect of Latin Hypercube constraint (the most used design in practice with the kriging) on maximin-optimal designs. This study shows that when the design has a small number of points, the addition of the Latin Hypercube constraint will be useful because it mitigates the drawbacks of maximin-optimal configurations (the position of the majority of points at the boundary of the input space). Following this study, an uniformity criterion called Radial discrepancy has been proposed in order to measure the uniformity of the points of the design according to their distance to the boundary of the input space. Then we show that the minimax-optimal design is the closest design to IMSE design (design which is adapted to prediction by kriging) but is also very difficult to evaluate. We then introduce a proxy for the minimax-optimal design based on the maximin-optimal design. Finally, we present an optimised implementation of the simulated annealing algorithm in order to find maximin-optimal designs. Our aim here is to minimize the probability of falling in a local minimum configuration of the simulated annealing. The second part of the thesis concerns a slightly different problem. If XN is space-filling-design of N points, there is no guarantee that any n points of XN (1 6 n 6 N) constitute a space-filling-design. In practice, however, we may have to stop the simulations before the full realization of design. The aim of this part is therefore to propose a new methodology to construct sequential of space-filling-designs (nested designs) of experiments Xn for any n between 1 and N that are all adapted to kriging prediction. We introduce a method to generate nested designs based on information criteria, particularly the Mutual Information criterion. This method ensures a good quality forall the designs generated, 1 6 n 6 N. A key difficulty of this method is that the time needed to generate a MI-sequential design in the highdimension case is very larg. To address this issue a particular implementation, which calculates the determinant of a given matrix by partitioning it into blocks. This implementation allows a significant reduction of the computational cost of MI-sequential designs, has been proposed.

Dans cette thèse, nous concentrons sur les sujets suivants: super-eulérien graphe, hamiltonien ligne graphes, le tolerant aux pannes hamiltonien laceabilité de Cayley graphe généré par des transposition arbres et plusieurs problèmes extrémaux concernant la (minimum et/ou maximum) taille des graphes qui ont la même propriété.Cette thèse comprend six chapitres. Le premier chapitre introduit des définitions et indique la conclusion des resultants principaux de cette thèse, et dans le dernier chapitre, nous introduisons la recherche de furture de la thèse. Les travaux principaux sont montrés dans les chapitres 2-5 comme suit:Dans le chapitre 2, nous explorons les conditions pour qu'un graphe soit super-eulérien.Dans la section 1, nous caractérisons des graphes dont le dégrée minimum est au moins de 2 et le nombre de matching est au plus de 3. Dans la section 2, nous prouvons que si pour tous les arcs xy∈E(G), d(x)+d(y)≥n-1-p(n), alors G est collapsible sauf quelques bien définis graphes qui ont la propriété p(n)=0 quand n est impair et p(n)=1 quand n est pair.Dans la section 3 de la Chapitre 2, nous trouvons les conditions suffisantes pour que un graphe de 3-arcs connectés soit pliable.Dans le chapitre 3, nous considérons surtout l'hamiltonien de 3-connecté ligne graphe.Dans la première section de Chapitre 3, nous montrons que chaque 3-connecté, essentiellement11-connecté ligne graphe est hamiltonien-connecté. Cela renforce le résultat dans [91]. Dans la seconde section de Chapitre 3, nous montrons que chaque 3-connecté, essentiellement 10-connecté ligne graphe est hamiltonien-connecté.Dans la troisième section de Chapitre 3, nous montrons que 3-connecté, essentiellement 4-connecté ligne graphe venant d'un graphe qui comprend au plus 9 sommets de degré 3 est hamiltonien. Dans le chapitre 4, nous montrons d'abord que pour tous $F\subseteq E(Cay(B:S_{n}))$, si $|F|\leq n-3$ et $n\geq 4$, il existe un hamiltonien graphe dans $Cay(B:S_{n})-F$ entre tous les paires de sommets qui sont dans les différents partite ensembles. De plus, nous renforçons le résultat figurant ci-dessus dans la seconde section montrant que $Cay(S_n,B)-F$ est bipancyclique si $Cay(S_n,B)$ n'est pas un star graphe, $n\geq 4$ et $|F|\leq n-3$.Dans le chapitre 5, nous considérons plusieurs problems extrémaux concernant la taille des graphes.Dans la section 1 de Chapitre 5, nous bornons la taille de sous-graphe provoqué par $m$ sommets de hypercubes ($n$-cubes). Dans la section 2 de Chapitre 5, nous étudions partiellement la taille minimale d'un graphe savant son degré minimum et son degré d'arc. Dans la section 3 de Chapitre 5, nous considérons la taille minimale des graphes satisfaisants la Ore-condition. / In this thesis, we focus on the following topics: supereulerian graphs, hamiltonian line graphs, fault-tolerant Hamiltonian laceability of Cayley graphs generated by transposition trees, and several extremal problems on the (minimum and/or maximum) size of graphs under a given graph property. The thesis includes six chapters. The first one is to introduce definitions and summary the main results of the thesis, and in the last chapter we introduce the furture research of the thesis. The main studies in Chapters 2 - 5 are as follows. In Chapter 2, we explore conditions for a graph to be supereulerian.In Section 1 of Chapter 2, we characterize the graphs with minimum degree at least 2 and matching number at most 3. By using the characterization, we strengthen the result in [93] and we also address a conjecture in the paper.In Section 2 of Chapter 2, we prove that if $d(x)+d(y)\geq n-1-p(n)$ for any edge $xy\in E(G)$, then $G$ is collapsible except for several special graphs, where $p(n)=0$ for $n$ even and $p(n)=1$ for $n$ odd. As a corollary, a characterization for graphs satisfying $d(x)+d(y)\geq n-1-p(n)$ for any edge $xy\in E(G)$ to be supereulerian is obtained. This result extends the result in [21].In Section 3 of Chapter 2, we focus on a conjecture posed by Chen and Lai [Conjecture~8.6 of [33]] that every 3-edge connected and essentially 6-edge connected graph is collapsible. We find a kind of sufficient conditions for a 3-edge connected graph to be collapsible.In Chapter 3, we mainly consider the hamiltonicity of 3-connected line graphs.In the first section of Chapter 3, we give several conditions for a line graph to be hamiltonian, especially we show that every 3-connected, essentially 11-connected line graph is hamilton- connected which strengthens the result in [91].In the second section of Chapter 3, we show that every 3-connected, essentially 10-connected line graph is hamiltonian-connected.In the third section of Chapter 3, we show that 3-connected, essentially 4-connected line graph of a graph with at most 9 vertices of degree 3 is hamiltonian. Moreover, if $G$ has 10 vertices of degree 3 and its line graph is not hamiltonian, then $G$ can be contractible to the Petersen graph.In Chapter 4, we consider edge fault-tolerant hamiltonicity of Cayley graphs generated by transposition trees. We first show that for any $F\subseteq E(Cay(B:S_{n}))$, if $|F|\leq n-3$ and $n\geq4$, then there exists a hamiltonian path in $Cay(B:S_{n})-F$ between every pair of vertices which are in different partite sets. Furthermore, we strengthen the above result in the second section by showing that $Cay(S_n,B)-F$ is bipancyclic if $Cay(S_n,B)$ is not a star graph, $n\geq4$ and $|F|\leq n-3$.In Chapter 5, we consider several extremal problems on the size of graphs.In Section 1 of Chapter 5, we bounds the size of the subgraph induced by $m$ vertices of hypercubes. We show that a subgraph induced by $m$ (denote $m$ by $\sum\limits_{i=0}^ {s}2^{t_i}$, $t_0=[\log_2m]$ and $t_i= [\log_2({m-\sum\limits_{r=0}^{i-1}2 ^{t_r}})]$ for $i\geq1$) vertices of an $n$-cube (hypercube) has at most $\sum\limits_{i=0}^{s}t_i2^{t_i-1} +\sum\limits_{i=0}^{s} i\cdot2^{t_i}$ edges. As its applications, we determine the $m$-extra edge-connectivity of hypercubes for $m\leq2^{[\frac{n}2]}$ and $g$-extra edge-connectivity of the folded hypercube for $g\leq n$.In Section 2 of Chapter 5, we partially study the minimum size of graphs with a given minimum degree and a given edge degree. As an application, we characterize some kinds of minimumrestricted edge connected graphs.In Section 3 of Chapter 5, we consider the minimum size of graphs satisfying Ore-condition.

In this thesis, we focus on the following topics: supereulerian graphs, hamiltonian line graphs, fault-tolerant Hamiltonian laceability of Cayley graphs generated by transposition trees, and several extremal problems on the (minimum and/or maximum) size of graphs under a given graph property. The thesis includes six chapters. The first one is to introduce definitions and summary the main results of the thesis, and in the last chapter we introduce the furture research of the thesis. The main studies in Chapters 2 - 5 are as follows. In Chapter 2, we explore conditions for a graph to be supereulerian.In Section 1 of Chapter 2, we characterize the graphs with minimum degree at least 2 and matching number at most 3. By using the characterization, we strengthen the result in [93] and we also address a conjecture in the paper.In Section 2 of Chapter 2, we prove that if $d(x)+d(y)\geq n-1-p(n)$ for any edge $xy\in E(G)$, then $G$ is collapsible except for several special graphs, where $p(n)=0$ for $n$ even and $p(n)=1$ for $n$ odd. As a corollary, a characterization for graphs satisfying $d(x)+d(y)\geq n-1-p(n)$ for any edge $xy\in E(G)$ to be supereulerian is obtained. This result extends the result in [21].In Section 3 of Chapter 2, we focus on a conjecture posed by Chen and Lai [Conjecture~8.6 of [33]] that every 3-edge connected and essentially 6-edge connected graph is collapsible. We find a kind of sufficient conditions for a 3-edge connected graph to be collapsible.In Chapter 3, we mainly consider the hamiltonicity of 3-connected line graphs.In the first section of Chapter 3, we give several conditions for a line graph to be hamiltonian, especially we show that every 3-connected, essentially 11-connected line graph is hamilton- connected which strengthens the result in [91].In the second section of Chapter 3, we show that every 3-connected, essentially 10-connected line graph is hamiltonian-connected.In the third section of Chapter 3, we show that 3-connected, essentially 4-connected line graph of a graph with at most 9 vertices of degree 3 is hamiltonian. Moreover, if $G$ has 10 vertices of degree 3 and its line graph is not hamiltonian, then $G$ can be contractible to the Petersen graph.In Chapter 4, we consider edge fault-tolerant hamiltonicity of Cayley graphs generated by transposition trees. We first show that for any $F\subseteq E(Cay(B:S_{n}))$, if $|F|\leq n-3$ and $n\geq4$, then there exists a hamiltonian path in $Cay(B:S_{n})-F$ between every pair of vertices which are in different partite sets. Furthermore, we strengthen the above result in the second section by showing that $Cay(S_n,B)-F$ is bipancyclic if $Cay(S_n,B)$ is not a star graph, $n\geq4$ and $|F|\leq n-3$.In Chapter 5, we consider several extremal problems on the size of graphs.In Section 1 of Chapter 5, we bounds the size of the subgraph induced by $m$ vertices of hypercubes. We show that a subgraph induced by $m$ (denote $m$ by $\sum\limits_{i=0}^ {s}2^{t_i}$, $t_0=[\log_2m]$ and $t_i= [\log_2({m-\sum\limits_{r=0}^{i-1}2 ^{t_r}})]$ for $i\geq1$) vertices of an $n$-cube (hypercube) has at most $\sum\limits_{i=0}^{s}t_i2^{t_i-1} +\sum\limits_{i=0}^{s} i\cdot2^{t_i}$ edges. As its applications, we determine the $m$-extra edge-connectivity of hypercubes for $m\leq2^{[\frac{n}2]}$ and $g$-extra edge-connectivity of the folded hypercube for $g\leq n$.In Section 2 of Chapter 5, we partially study the minimum size of graphs with a given minimum degree and a given edge degree. As an application, we characterize some kinds of minimumrestricted edge connected graphs.In Section 3 of Chapter 5, we consider the minimum size of graphs satisfying Ore-condition.

In this thesis, we propose novel optimization and spatial queueing models that expand the currently existing methods by allowing multiple servers to be located at the same station and multiple servers to be dispatched to a single call. In particular, a mixed integer linear programming (MILP) model is introduced that determines how to locate and dispatch ambulances such that the coverage level is maximized. The model allows multiple servers to be located at the same station and balances the workload among them while maintaining contiguous first priority response districts. We also propose an extension to the approximate Hypercube queueing model by allowing multi-server dispatches. Computational results suggest that both models are effective in optimizing and analyzing the emergency systems. We also introduce the M[G]/M/s/s queueing model as an extension to the M/M/s/s model which allows for multiple servers to be assigned to a single customer.

En este trabajo se propone y desarrolla una topología en k-hipercubos que resuelve los principales inconvenientes asociados a la topología en hipercubo convencional. Los resultados obtenidos son muy prometedores, con aplicaciones tanto en el campo de la voz sobre IP, como en muchos otros campos que precisen de un intercambio de información muchos a muchos. Sobre la topología propuesta se define el protocolo Darkcube, que es una propuesta de protocolo totalmente distribuido basado en el concepto de darknet, posibilitando la realización de conversaciones muchos a muchos incluyendo audio, vídeo, texto y datos de geoposicionamiento, entre otros. También se propone un método de codificación de coordenadas de geoposicionamiento que resulta especialmente eficiente en el aprovechamiento del ancho de banda sobrante en las comunicaciones muchos a muchos que proporciona Darkcube. Durante el desarrollo de este trabajo, se ha implementado el simulador DarkcubeEmu; herramienta que posibilita la obtención de resultados relevantes en términos de la calidad de la comunicación. Finalmente, utilizando como base el protocolo Darkcube, se propone un protocolo de seguridad que traslada un esquema de infraestructura de clave pública a un protocolo totalmente distribuido, como es Darkcube; garantizando, de esta forma, la confidencialidad en las comunicaciones y la legitimidad de la identidad asociada a cada uno de sus miembros.

Étude de divers problèmes lies aux graphes de Hamming

The Turán problem is a fundamental problem in extremal graph theory. It asks what the maximum number of edges a given graph G can have, not containing some forbidden graph H, and is solved using the Turán number ex(n,H), density π(H) and graph Tr(n). Turán's theorem tells us that the Turán graph Tr(n) is the largest Kr+1-free simple graph on n vertices. This paper is an overview of Turán problems for cliques Kn, hypercubes Qn and Hamming graphs H(s,d). We end it by proving a new result we call "the layer theorem", solving the Hamming-Turán problem using a method of creating layers of vertices in a graph. This theorem gives a lower bound for the Hamming-relative Turán density as follows: <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?%5Cpi_%7Bs,d%7D(%5Cmathcal%7BH%7D_%7Bs,d%7D,F)%20%5Cgeq%201%20-%20%5Cdfrac%7Bf+g%7D%7B%7C%7CH(s,d)%7C%7C%7D" /> where <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?f%20=%20%5Cbinom%7Bs%7D%7B2%7D%5Cleft(1-%5Cdfrac%7Br-2%7D%7Br-1%7D%5Cright)ds%5E%7Bd-1%7D%20%5Ctext%7B%20and%20%7D%20g%20=%20%5Csum_%7Bi=1%7D%5E%7Bn/(t-1)%7D%20(d-i(t-1))(s-1)%5E%7Bi(t-1)+1%7D%5Cbinom%7Bd%7D%7Bi(t-1)%7D" /> for the forbidden graph F stretching over t layers and r = χ(F). / Turán-problemet är det fundamentala problemet inom extremal grafteori. Det ställer frågan vad det maximala antalet kanter en given graf G kan ha utan att innehålla någon förbjuden graf H, och löses med hjälp av Turán-talet ex(n,H), -densiteten π(H) and -grafen Tr(n). Turáns sats säger oss att Turán-grafen Tr(n) är den största Kr+1-fria enkla grafen på n hörn. Denna uppsats är en överblick av Turán-problem i klickar Kn, hyperkuber Qn och Hamming-grafer H(s,d). Vi avslutar den med att bevisa ett nytt resultat som vi kallar "lagersatsen", vilket löser Hamming-Turán-problemet med hjälp av en metod som skapar lager av hörnen i en graf. Lagersatsen ger en undre gräns för den Hamming-relativa Turán-densiteten enligt följande: <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?%5Cpi_%7Bs,d%7D(%5Cmathcal%7BH%7D_%7Bs,d%7D,F)%20%5Cgeq%201%20-%20%5Cdfrac%7Bf+g%7D%7B%7C%7CH(s,d)%7C%7C%7D" /> där <img src="http://www.diva-portal.org/cgi-bin/mimetex.cgi?f%20=%20%5Cbinom%7Bs%7D%7B2%7D%5Cleft(1-%5Cdfrac%7Br-2%7D%7Br-1%7D%5Cright)ds%5E%7Bd-1%7D%20%5Ctext%7B%20and%20%7D%20g%20=%20%5Csum_%7Bi=1%7D%5E%7Bn/(t-1)%7D%20(d-i(t-1))(s-1)%5E%7Bi(t-1)+1%7D%5Cbinom%7Bd%7D%7Bi(t-1)%7D" /> för den förbjudna grafen F som sträcker sig över t lager samt r = χ(F).

Digital soil mapping (DSM) relies on quantitative relationships between easily measured environmental covariates and field and laboratory data. We applied innovative sampling and inference techniques to predict the distribution of soil attributes, taxonomic classes, and dominant vegetation across a 30,000-ha complex Great Basin landscape in southwestern Utah. This arid rangeland was characterized by rugged topography, diverse vegetation, and intricate geology. Environmental covariates calculated from digital elevation models (DEM) and spectral satellite data were used to represent factors controlling soil development and distribution. We investigated optimal sample size and sampled the environmental covariates using conditioned Latin Hypercube Sampling (cLHS). We demonstrated that cLHS, a type of stratified random sampling, closely approximated the full range of variability of environmental covariates in feature and geographic space with small sample sizes. Site and soil data were collected at 300 locations identified by cLHS. Random forests was used to generate spatial predictions and associated probabilities of site and soil characteristics. Balanced random forests and balanced and weighted random forests were investigated for their use in producing an overall soil map. Overall and class errors (referred to as out-of-bag [OOB] error) were within acceptable levels. Quantitative covariate importance was useful in determining what factors were important for soil distribution. Random forest spatial predictions were evaluated based on the conceptual framework developed during field sampling.