Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The wounds measurement is an important factor in the evaluation healing because it provides parameters that indicate the progress of wound healing. The objective is to compare the area measured of venous ulcers through AutoCAD® and Image Tool software. The assessment of wound measurement methods is constituted as a gap in the knowledge produced by Brazilian nursing and as one of the fundamental aspects of the wounds evaluation. It is an evaluative analysis of reproducibility tests with a quantitative approach. The study included patients with venous ulcers treated at an outpatient clinic during the data collection period (March-July 2015), higher the age of 18 years with cognitive and verbal skills preserved, which had ulcers covering one side of the lower limb (n=21). For the data collection procedure, nurses that participated of the Skin Injury Study Group were contributors of research, being called evaluators. Each patient had his ulcer photographed by the researcher and by an evaluator with Cam Fujufilm Finepix S 14 Mega pixels. The photographs were transferred to the personal researcher's notebook, both the evaluator and the researcher carried out the measurements of the two photographs, with the two softwares. For the analysis of data mean and standard error, Wilcoxon test, intraclass correlation coefficient and concordance correlation coefficient were used, in addition to Bland and Altman procedure. The ethical precepts were based on Resolution No. 466/2012. 21 individuals participated in the study, who presented a total of 36 venous ulcers (72 photos). The average age of participants was 60.9, the majority of subjects were male, had one venous ulcer, with 3,0 years median time of existence of injury. The size of the ulcers had great amplitude. There was difference between the measurements of the researcher and the evaluators in AutoCAD® software, with no difference in the Image tool. However, the values of the correlation coefficients were strong in both cases. There was no significant difference between the measurements in both softwares, there is strong intraclass correlation and agreement between both, which seem to be more accurate in measuring wounds with an area> 10 cm². Measurements in AutoCAD and Image Tool softwares feature intraclass correlation and strong concordance, so they are reproducible. Still they appears to be more accurate when used to measure wound area> 10 cm². Therefore, the use of both is suitable for the measurement of venous ulcers with an area> 10 cm². / A mensuração de feridas é um dado importante no processo de avaliação da cicatrização porque fornece parâmetros que indicam a evolução da cicatrização da ferida. O objetivo é comparar a mensuração de área de úlceras venosas por meio dos softwares AutoCAD® e Image Tool. A avaliação de métodos de mensuração de feridas constitui-se como uma lacuna no conhecimento produzido pela enfermagem brasileira e como um dos aspectos fundamentais da avaliação de feridas. Trata-se de uma pesquisa de avaliação de reprodutibilidade de testes, com abordagem quantitativa. Participaram do estudo os pacientes portadores de úlceras venosas, atendidos em um ambulatório no período de coleta de dados (março a julho de 2015), com idade superior a 18 anos, com capacidade cognitiva e verbal preservadas, que possuíam úlceras cobrindo uma face do membro inferior (n=21). Para o procedimento de coleta de dados, os enfermeiros participantes do Grupo de Estudos de Lesões de Pele foram colaboradores da pesquisa, sendo denominados de avaliadores. Cada paciente teve sua úlcera fotografada pelo pesquisador e por um avaliador com Câmera Fujufilm Finepix S 14 Mega pixels. As fotografias foram transferidas para o notebook pessoal do pesquisador, tanto o avaliador quanto o pesquisador realizaram as mensurações das duas fotografias, com os dois softwares. Para a análise dos dados foram utilizados média e erro padrão, teste de Wilcoxon, Coeficiente de Correlação Intraclasse e Coeficiente de Correlação de Concordância, além do procedimento de Bland e Altman. Os preceitos éticos foram baseados na Resolução nº 466/2012. Participaram do estudo 21 indivíduos, que apresentaram, no total, 36 úlceras venosas (72 fotografias). A idade média dos participantes foi de 60,9, a maioria dos indivíduos era do sexo masculino, apresentava uma úlcera venosa, com tempo mediano de existência da lesão de 3,0 anos. Os tamanhos das úlceras apresentaram grande amplitude. Observou-se diferença entre as mensurações da pesquisadora e dos avaliadores no software AutoCAD®, sem diferença no Image tool. Apesar disso, os valores dos coeficientes de correlação foram fortes em ambos os casos. Não foi observada diferença significativa entre as mensurações nos dois softwares, existe forte correlação intraclasse e de concordância entre os ambos, os quais parecem ser mais precisos na mensuração de feridas com área > 10 cm². As mensurações nos softwares AutoCAD® e Image Tool apresentam correlação intraclasse e concordância fortes, ou seja, são reprodutíveis. Ainda, parecem ser mais precisos quando utilizados para mensurar feridas com área > 10 cm². Portanto, o uso de ambos é indicado para a mensuração de úlceras venosas com área > 10 cm².

The common topic of this thesis is boundedness of integral and supremal operators between weighted function spaces. The first type of results are characterizations of boundedness of a convolution-type operator between general weighted Lorentz spaces. Weighted Young-type convolution inequalities are obtained and an optimality property of involved domain spaces is proved. Additional provided information includes an overview of basic properties of some new function spaces appearing in the proven inequalities. In the next part, product-based bilinear and multilinear Hardy-type operators are investigated. It is characterized when a bilinear Hardy operator inequality holds either for all nonnegative or all nonnegative and nonincreasing functions on the real semiaxis. The proof technique is based on a reduction of the bilinear problems to linear ones to which known weighted inequalities are applicable. Further objects of study are iterated supremal and integral Hardy operators, a basic Hardy operator with a kernel and applications of these to more complicated weighted problems and embeddings of generalized Lorentz spaces. Several open problems related to missing cases of parameters are solved, thus completing the theory of the involved fundamental Hardy-type operators. / Operators acting on function spaces are classical subjects of study in functional analysis. This thesis contributes to the research on this topic, focusing particularly on integral and supremal operators and weighted function spaces. Proving boundedness conditions of a convolution-type operator between weighted Lorentz spaces is the first type of a problem investigated here. The results have a form of weighted Young-type convolution inequalities, addressing also optimality properties of involved domain spaces. In addition to that, the outcome includes an overview of basic properties of some new function spaces appearing in the proven inequalities.  Product-based bilinear and multilinear Hardy-type operators are another matter of focus. It is characterized when a bilinear Hardy operator inequality holds either for all nonnegative or all nonnegative and nonincreasing functions on the real semiaxis. The proof technique is based on a reduction of the bilinear problems to linear ones to which known weighted inequalities are applicable.  The last part of the presented work concerns iterated supremal and integral Hardy operators, a basic Hardy operator with a kernel and applications of these to more complicated weighted problems and embeddings of generalized Lorentz spaces. Several open problems related to missing cases of parameters are solved, completing the theory of the involved fundamental Hardy-type operators. / <p>Artikel 9 publicerad i avhandlingen som manuskript med samma titel.</p>

Les entrées des projections de dLGN et FB des aires corticales à L1 du V1 de la souris sont discontinues. Elles correspondent à un motif d'expression M2AChR. Ce motif est aussi observé chez le rat et le singe. Les neurones en L2/3 alignés avec les zones M2+ ont une grande acuité spatiale, tandis que dans les zones M2- ont une grande acuité temporelle. Ensemble, les zones M2 + et M2- forment des domaines constants. Ils codent des sous-régions du RF, de sorte que plusieurs domaines contribuent à l'image d'un point du champ visuel.En utilisant des traceurs rétrogrades, nous montrons un principe d'organisation générale fondée sur une règle de la distance exponentielle (EDR) et la géométrie corticale. Nous trouvons des invariants de réseau, mais aussi des différences significatives, telles que des connexions de longue distance beaucoup plus faibles chez le macaque. Une EDR est aussi présente à l'échelle locale, à moins de 1,5 mm, ce qui indique qu'elle pourrait être une propriété universellement applicable à toutes les échelles et chez toutes les mammifères.41 injections avec des traceurs rétrogrades ont été faites dans 22 des 45 régions du néocortex de la souris. Nous avons aplati le cortex et utilisé des critères histologiques et génétiques pour la répartition des neurones marqués dans les aires corticales. Pour chaque connexion, un poids a été déterminé. La cohérence entre les animaux est influencée par le poids moyen et la taille de l'injection. La distribution lognormale des connexions à une aire corticale couvre 5 ordres de grandeur et constitue un profil de connectivité qui est caractéristique de chaque aire. La matrice cortico-corticale présente une densité de 96% / We find that inputs to the non-columnar mouse V1 from the dLGN and FB projections from cortical areas to L1 are patchy. The patches are matched to a pattern of M2AChR expression at ?xed locations of mouse, rat, and monkey V1. Neurons in L2/3 aligned with M2-rich patches have high spatial acuity, whereas cells in M2-poor zones have high temporal acuity. Together M2+ and M2-zones form constant-size domains that are repeated across V1. Domains map subregions of the RF, such that multiple copies are contained within the point image. Using tract tracing data from macaque and mouse, we show a general organizational principle based on an exponential distance rule (EDR) and cortical geometry. We find network invariants between mouse and macaque, but also significant differences, such as fractionally smaller and much weaker long distance connections in the macaque than in mouse. An EDR holds at local scales as well (within 1.5 mm), indicating that it might be a universally valid property across all scales and across the mammalian class.41 injections with retrograde tracers were made in 22 of the 40 areas of the mouse neocortex. Flat mounts of the cortex complete with comprehensive histological and genetic criteria enabled allocation of counts of labeled neurons to individual cortical areas. A weight was determined for each connection. Consistency across animals was systematically influenced by mean weight and injection size. The lognormal distribution of connections to a cortical area spanned 5 orders of magnitude and constituted a connectivity profile that was highly characteristic for each area. The resulting matrix showed that 96% of connections that can exist do exist

Le travail effectué au cours de cette thèse se divise en trois. Pour la première partie, 12 injections de traceurs rétrogrades ont été effectuées chez le macaque cynomolgus à différentes excentricités visuelles de V1, V2, V4 et MT. Nous observons que, si les injections effectuées au sein des territoires correspondant aux représentations centrales (i.e fovéale) de l'hémi-champs visuel reçoivent préférentiellement des projections issues de la voie ventrale, les injections placées au sein des représentations périphériques marques préférentiellement les neurones d'aires connues pour faire partie de la voie dorsale. Les injections excentriques affichent une décroissance exponentielle des poids de connexion avec l'augmentation des distances, comme décrit précédemment (Markov et al. 2014). Ceci plaide en faveur de l'hypothèse que la règle exponentielle de distance (EDR en anglais) (Ercsey-Ravasz et al 2013) est assez puissante pour expliquer la formation des voies fonctionnelles du cortex visuel. De plus, cela suggère fortement que l'EDR s'applique à n'importe quel point du manteau cortical, indépendamment de toutes limites d'aires, et que, par conséquent, le cortex est mieux décrit comme une gradient continu de propriétés progressivement changeantes, au lieu d'une mosaïque de aires corticales bien définies. De plus, cela a de profondes implications concernant l'évolution biologique des aires corticales ainsi que du cortex dans son ensemble. En seconde partie, tractographie et traçage de voies furent comparés, via l'utilisation d'IRM de diffusion. Une analyse en tractographie fut effectuée, et la matrice de connectivité résultante comparée à celle obtenue par traçage de voies (Markov et al., 2014). Les résultats préliminaires de cette thèse tombent dans le même intervalle de succès de détection que les études précédentes : entre 70 et 77 % de la connectivité évaluée par tractographie est correcte (Azadbakht et al 2015, Calabrese et al 2015a, Donahue et al 2016). Cela met également en évidence un problème de spécificité des connexions détectées par la tractographie, qui ne parvient pas à reconnaître les connexions absentes avec précision, un problème particulièrement visible dans le contexte des aires visuelles de bas niveau, chez qui la connectivité est connue pour être très spécifique (Tigges et al 1973, Tigges et al 1974, Martinez-Millan & Hollander 1975, Kaas & Lin 1977, Wong-Riley 1978). En conclusion, les propriétés de connectivité du cortex ne devraient pas, à ce stade, être déduites la tractographie basée sur l'imagerie de diffusion. Pour la dernière partie, 6 injections ont été réalisées chez 3 animaux bilatéralement énucléés (3 dans le « Default Extrastriate Cortex » - DEC ; 1 dans V2 et 2 dans V4) afin d'évaluer comment la connectivité visuelle est affectée par la perte des afférences rétiniennes au thalamus. Comme rapportées antérieurement, des gyri anormaux sont observés sur l'operculum occipital, dont le territoire est dominé par la DEC, un cortex hybride entre cortex strié et extrastrié (Dehay et al 1996a, Dehay et al 1996b). De plus, l'anatomie du LGN est massivement perturbée. Les afférences en provenance du pulvinar et du LGN sont éparses et mal ciblées lors d'injections placées côte-à-côte. La connection du pulvinar au DEC est d'un ordre de grandeur supérieure à une connexion normale du pulvinar à V1, alors que la même comparaison pour le LGN est d'un ordre de grandeur inférieure. Ces changements en matière de connectivité s'amenuisent en allant plus haut au sein de la hiérarchie visuelle corticale. En effet, les changements les plus frappants se produisent dans V1 (aberrant gyri, DEC, ventralisation, etc.) alors que V4 semble être peu perturbé par l'absence d'afférences rétiniennes. Du point de vue du développement du système visuel, cela plaide pour un effet localisé des entrées thalamiques (i.e. seulement V1 et V2) ainsi que pour une absence d'effet domino en cascade compromettant l'ensemble du cerveau / The work performed during this thesis is three fold. In the first part, 12 injections of retrograde tracers were performed in cynomolgus macaques at different visual eccentricities in V1, V2, V4 and MT. We observe that, while injections performed in territories corresponding to the central (i.e. foveal) representation of the visual hemifield preferentially receive inputs from areas belonging to the ventral pathway, injections placed in peripheral representations preferentially labels neurons in areas known to be part of the dorsal pathway. Eccentric injections return an exponential decay of connexion strength with increasing distances, as previously described (Markov et al, 2014). This strongly argues in favour of the Exponential Distance Rule (EDR) being powerful enough to explain the formation of the functional streams of the visual cortex. Firstly, it strongly suggests that the EDR is applicable at any point of the cortical sheet regardless of area boundaries, and that therefore the cortex is better seen as a continuous sheet of gradually changing properties, instead of a mosaic of well-defined cortical areas. Secondly, it helps inform and refine our current definition of an area towards a more comprehensive one which includes topological location. Finally, it has deep implications regarding the evolution of cortical areas as well as the cortex as a whole. In the second part, a comparison was attempted between tractography and tract tracing, with the high definition diffusion MRI scan of one macaque brain. Tractography was computed on it, and the resulting connectivity matrix compared to tract tracing (Markov et al, 2014). The preliminary results obtained in this thesis fall in the same range of detection success than previous studies: between 70 and 77 % of the connectivity assessed by dMRI-based tractography is correct using the same Receiver Operating Charateristic (ROC) approach (Azadbakht et al 2015, Calabrese et al 2015a, Donahue et al 2016). It also highlights a problem of specificity of connexions detected by tractography. It fails to accurately discriminate empty connexions (i.e. absent) as empty, a problem especially salient in the context of early visual areas, the connectivity of which is known to be very specific (Tigges et al 1973, Tigges et al 1974, Martinez-Millan & Hollander 1975, Kaas & Lin 1977, Wong-Riley 1978). Overall, caution is for now warranted and connectivity properties should not at this point be inferred on diffusion MRI based tractography. In the last part, 6 injections were performed in 3 bilaterally enucleated animals (3 in the Default Extrastriate Cortex - DEC, 1 in V2 and 2 in V4) in order to assess how the visual connectivity is impacted by the loss of retinal inputs to the thalamus. As previously reported, abnormal gyri are observed on the operculum, whose territory is now dominated by the DEC, a hybrid cortex combining histological and histochemical feature of striate and exstrastiate cortex (Dehay et al 1996a, Dehay et al 1996b). Also, the LGN anatomy is massively perturbed, losing its heterogeneity of strictly defined layers for a homogeneous structure. Inputs from the pulvinar and the LGN were found to be unfocused, scattered and unsegregated in side-by-side injections. The pulvinar input to the DEC was found to be an order of magnitude higher than a normal pulvinar-to-V1 connexion, when the same comparison for the LGN was found to be an order of magnitude lower. These changes in connectivity appear to diminish in importance as one goes up the visual hierarchy. Indeed the most striking changes are in V1 (aberrant gyri, DEC, ventralisation, etc.) whereas V4 appears to be seemingly unperturbed by the absence of retinal inputs. Regarding the development of the visual system, this fact argues for a localised effect of the thalamic inputs (i.e. only V1 and V2) and an absence of domino-effect cascade that would perturb the entirety of the brain, despite the quite drastic enucleation process

Les systèmes implicites (dits aussi « descripteurs ») peuvent décrire des processus régis à la fois par des équations dynamiques et statiques et permettent de préserver la structure des systèmes physiques. Ils comportent trois types de modes : dynamiques finis, infinis (réponse temporelle impulsive (en cas continu) ou acausale (en cas discret)) et statiques. Dans le cadre du formalisme descripteur, les contributions de cette thèse sont triples : i) revisiter des résultats existants pour les systèmes d’état, ii) étendre certains résultats classiques au cas des systèmes implicites, iii) résoudre rigoureusement des problèmes de commande non standard. Ainsi, le présent mémoire commence par revisiter les résultats concernant la caractérisation LMI stricte de la dissipativité, les caractérisations de l’admissibilité et des performances H2 ou H∞ par LMI étendues et les équations de Sylvester et de Riccati généralisées. Il aborde dans un deuxième temps, le problème de stabilisation simultanée, avec ou sans critère H∞, à travers l’extension de certains résultats récents au cas des systèmes implicites. La solution proposée s’appuie sur la résolution combinée d’une équation algébrique de Riccati généralisée (GARE) et d’un problème de faisabilité sous contrainte LMI stricte. Il traite enfin des problèmes H2 et H∞ non standards : i) en présence de pondérations instables voire impropres, ii) sous contraintes de régulation; dans le cas des systèmes implicites. Ces dernières contributions permettent désormais de traiter rigoureusement, sans approximations ou transformations, de nombreux problèmes H2 ou H∞ formalisant des problèmes pratiques de commande, dont ceux faisant intervenir une pénalisation haute fréquence de la commande ou un modèle interne instable des signaux exogènes. / The descriptor systems have been attracting the attention of many researchers over recent decades due to their capacity to preserve the structure of physical systems and to describe static constraints and impulsive behaviors. Within the descriptor framework, the contributions of this dissertation are threefold: i) review of existing results for state-space systems, ii) generalization of classical results to descriptor systems, iii) exact and analytical solutions to non standard control problems. A realization independent Kalman-Yakubovich-Popov (KYP) lemma and dilated LMI characterizations are deduced for descriptor systems. The solvability and corresponding numerical algorithms of generalized Sylvester equations and generalized algebraic Riccati equations (GARE) associated with descriptor systems are provided. In addition, the simultaneous H∞ control problem is considered through extending recently reported results. A sufficient condition is proposed through a combination of a generalized algebraic Riccati equation and a set of LMIs. Moreover, the nonstandard H2 and H∞ control problems with unstable and/or nonproper weighting functions or subject to regulation constraints are addressed. These contributions allow, without approximation or transformation, dealing with many practical problems defined within H2 or H∞ control methodologies, where the control signals are penalized at high frequency or unstable internal models specified by external signals is involved.

Thesis (M. Sc. Agriculture (Animal Production)) -- University of Limpopo, 2019 / Two experiments were conducted to determine the effect of dietary methionine level on productivity and carcass characteristics of Ross 308 broiler chickens. In each experiment, the diets were isocaloric and isonitrogenous but with different dietary methionine levels. Five diets were formulated to contain dietary methionine levels of 4, 5, 6, 8 or 9g/kg DM. The first experiment commenced with 300 unsexed Ross 308 broiler chickens with initial average live weights of 42 ± 2g per chicken. The chickens were randomly assigned to five treatments with five replications, resulting in 25-floor pens with 12 chickens per replicate. The second experiment commenced with 150 male Ross 308 broiler chickens with initial average live weight of 637 ± 12g per chicken. The chickens were randomly assigned to five treatments with three replications, resulting in 15-floor pens with 10 chickens per replicate. A complete randomized design was used in each experiment. Data was analysed using the General Linear Model (GLM) procedures of the statistical analysis of variance, Version 9.3.1 software program. Where there were significant differences, mean separation was done using the Tukey test at the 5% level of significance. A quadratic regression model was used to determine the optimal productivity of the chickens while a linear model was used to determine the relationships between dietary methionine level and responses by the chickens in the variables measured. The treatments for the first experiment were UM4 (4g methionine/kg DM), UM5 (5g methionine/kg DM), UM6 (6g methionine/kg DM), UM8 (8g methionine/kg DM) and UM9 (9g methionine/kg DM). Feed intake, growth rate, feed conversion ratio (FCR), metabolisable energy intake and nitrogen retention of unsexed Ross 308 broiler chickens aged one to 21 days were not affected (P>0.05) by dietary methionine level. Similarly, dietary methionine level did not have any effect (P>0.05) on diet crude protein (CP), acid detergent fibre (ADF), neutral detergent fibre (NDF) and fat digestibilities in unsexed Ross 308 broiler chickens aged 14 to 21 days. Dietary methionine level did not have any effect on live weights of broiler chickens at 21 days. Live weights of unsexed Ross 308 broiler chickens aged 7 or 14 days were not improved (P>0.05) by increasing dietary methionine level from 4 to 9g/kg DM. Crop, gizzard and small intestine weights and crop, proventriculus and gizzard digesta pH values of unsexed Ross 308 broiler chickens aged 21 days were not affected v (P>0.05) by dietary methionine level. Similarly, dietary methionine level did not improve (P>0.05) caecum and large intestine lengths of unsexed Ross 308 broiler chickens aged 21 days. However, dietary methionine level affected (P<0.05) dry matter (DM) and ash digestibilities of unsexed Ross 308 broiler chickens aged 14 to 21 days. Proventriculus and large intestine weights, gastrointestinal tract and small intestine lengths of unsexed Ross 308 broiler chickens aged 21 days were improved (P<0.05) by increasing dietary methionine level. In addition, increasing dietary methionine level increased (P<0.05) small and large intestine digesta pH values of broiler chickens aged 21 days. Thus, dry matter digestibility, live weights at day 7 ad 14, caecum length, large intestine length and digesta pH were optimized at different dietary methionine levels of 7.26, 5.29, 4.99, 6.80, 4.84 and 6.37g/kg DM feed, respectively. The treatments for the second experiment were MM4 (4g methionine/kg DM), MM5 (5g methionine/kg DM), MM6 (6g methionine/kg DM), MM8 (8g methionine/kg DM) and MM9 (9g methionine/kg DM). Dietary methionine level did not have effect (P>0.05) on feed intake of male Ross 308 broiler chickens aged 6 weeks. However, dietary methionine level improved (P<0.05) feed intake of male Ross 308 broiler chickens aged 4 or 5 weeks. Live weights of male Ross 308 broiler chickens aged 28 days were not affected (P>0.05) by dietary methionine level. However, live weights of male Ross 308 broiler chickens aged 35 and 42 days were affected (P<0.05) by dietary methionine level. Similarly, dietary methionine level affected (P<0.05) DM, CP, ADF, NDF, fat and ash digestibilities of male Ross 308 broiler chickens aged 35 to 42 days. Thus, dietary methionine levels of 6.93, 7.70, 6.85 and 11.27g/kg DM optimized dry matter, CP and fat digestibilities, and live weight of male broiler chickens aged 42 days. Dietary methionine level did not affect (P>0.05) FCR, growth rate and metabolisable energy intakes of male Ross 308 broiler chickens. Increasing dietary methionine level from 4 to 9g/kg DM improved (P<0.05) nitrogen retention of male Ross 308 broiler chickens aged 22 to 42 days. Dietary methionine level did not have any effect (P>0.05) on proventriculus, gizzard, caecum and large intestine weights, caecum, small and large intestine lengths, and crop, gizzard, caecum and large intestine digesta pH values of male Ross 308 broiler chickens aged 42 days. Crop and small vi intestine weights and gastrointestinal tract lengths of male Ross 308 broiler chickens aged 42 days were improved (P<0.05) by dietary methionine level. Similarly, dietary methionine level affected (P<0.05) proventriculus and small intestine digesta pH values of male Ross 308 broiler chickens aged 42 days. Thus, dietary methionine levels of 6.558 and 7.851g/kg DM optimized broiler chicken crop weight and GIT length. Dietary methionine level affected (P<0.05) carcass organ weights of male Ross 308 broiler chickens aged 42 days. Increasing dietary methionine level increased chicken breast meat weight. However, there was no clear trend for the other carcass organs. Meat flavour and shear force values of male Ross 308 broiler chickens aged 42 days were not affected (P>0.05) by dietary methionine level. However, dietary methionine level affected meat tenderness and juiciness. Thus, dietary methionine levels of 10.09 and 13.32g/kg DM optimized broiler chicken meat tenderness and juiciness. . / National Research Foundation (NRF) and VLIROUS

The work deals with the safety of technological systems and can be divided into two main thematic units. The attention in the first theoretic unit is paid to aspects that affect the safety of technological systems during their technical life. It describes, together with elementary norms and acts solving the products safety, the most important principles for creating and analysing the safety. It clarifies the origin of company’s risk management, part of its quality management, mentions the faults in which products can be found and also describes the failures which have the biggest effect on these faults and of course on the safety too. Attention is further paid to the reliability analysis, as the tool that helps to create the products safety in the pre-manufacturing phase and also to technical diagnostic, as the tool to keep the safety in after-manufacturing phase. The second practical unit arithmetically analyses the safety of two high-voltage transformers. For that it was used the database of real operation numerical values that were gained during diagnostic investigations on these transformers. In addition it describes mentioned diagnostics methods and gives the new procedure for analyse of transformers’ safety by criteria matrix and weighted sum approach. The main contribution of the work can be considered the checkout and the evaluation of this procedure for use in elaboration of high-voltage transformers diagnostic values for analyzing their safety.

Tato zpráva popisuje  experimentální práci na statistické jazykovém modelování pomocí rekurentních neuronových sítí (RNN). Je zde předložen důkladný přehled dosud publikovaných prací, následovaný popisem algoritmů pro trénování příslušných modelů. Většina z popsaných technik byla implementována ve vlastním nástroji, založeném na knihovně Theano. Byla provedena rozsáhlá sada experimentů s modelem Jednoduché rekurentní sítě (SRN), která odhalila některé jejich dosud nepublikované vlastnosti. Při statické evaluaci modelu byly dosažené výsledky relativně cca. o 2.7 % horší, než nejlepší publikované výsledky. V případě dynamické evaluace však bylo dosaženo relativního zlepšení o 1 %. Dále bylo experimentováno i s modelem Strukturně omezené rekurentní sítě, ale ten se nepodařilo natrénovat k předpokládáným výkonům. Konečně bylo navrženo rozšíření SRN, pojmenované Náhodně prořidlá rekurentní neuronová síť. Experimentálně bylo potvrzeno, že RS-RNN dosahuje lepších výsledků v učení vlastního trénovacího korpusu a kombinace několika RS-RNN modelů přináší o 30 % větší zlepšení než kombinace stejného počtu SRN.

Die vorliegende Arbeit befasst sich mit einer Vermutung von Sarnak aus dem Jahre 2010 über die Orthogonalität von durch deterministische dynamische Systeme induzierte Folgen zur Möbiusschen μ-Funktion. Ihre Hauptresultate sind zum einen der Ergodensatz mit Möbiusgewichten, welcher eine maßtheoretische (schwächere) Version von Sarnaks Vermutung darstellt, und zum anderen die bereits gesicherte Gültigkeit der genannten Vermutung in Spezialfällen, wobei hier exemplarisch unter anderem der Thue–Morse Shift und Schiefprodukterweiterungen von rationalen Rotationen auf dem Kreis gewählt worden sind. Zum Zwecke der Motivation zeigen wir, dass eine gewisse Wachstumsabschätzung für die Mertensfunktion äquivalent ist zum Primzahlsatz und skizzieren ein Resultat, welches die Äquivalenz einer weiteren solchen Abschätzung zur Riemannschen Vermutung liefert, um auf diese Weise die Bedeutung der Möbiusfunktion für die Zahlentheorie herauszustellen. Da sie für das Verständnis von Sarnaks Vermutung unerlässlich ist, geben wir eine Einführung in die Theorie der Entropie dynamischer Systeme auf Grundlage der Definitionen von Adler–Konheim–McAndrew, Bowen–Dinaburg und Kolmogorov–Sinai. Ferner berechnen wir die topologische Entropie des Thue–Morse Shifts und von Schiefprodukterweiterungen von Rotatione auf dem Kreis. Wir studieren die ergodische Zerlegung T-invarianter Maße auf kompakten metrischen Räumen mit stetiger Transformation T, welche wir für den Beweis des Ergodensatzes mit Möbiusgewichten benötigen. Sodann beweisen wir den genannten gewichteten Ergodensatz. Wir geben eine hinreichende Bedingung an für das Erfülltsein von Sarnaks Vermutung in einem gegebenen dynamischen System, welche im anschließenden Kapitel Anwendung findet. So wird nachgewiesen, dass Sarnaks Vermutung im Falle des Thue–Morse Shifts und von Schiefprodukterweiterungen von rationalen Rotationen auf dem Kreis erfüllt ist. Abschließend wird gezeigt, dass Sarnaks Vermutung sich als Konsequenz aus einer Vermutung von Chowla ergibt. / The thesis in hand deals with a conjecture of Sarnak from 2010 about the orthogonality of sequences induced by deterministic dynamical systems to the Möbius μ-function. Its main results are the ergodic theorem with Möbius weights, which is a measure theoretic (weaker) version of Sarnak’s conjecture, and the already assured validity of Sarnak’s conjecture in special cases, where we have exemplarily chosen the Thue–Morse shift and skew product extensions of rational rotations on the significance of the Möbius function for number theory. Since it is essential for the understanding of Sarnak’s conjecture we give an introduction to the theory of entropy of dynamical systems based on the definitions of Adler–Konheim–McAndrew, Bowen–Dinaburg and Kolmogorov–Sinai. Furthermore, we calculate the topological entropy of the Thue–Morse shift and of skew product extensions of rotations on the circle. We study the ergodic decomposition for T-invariant measures on compact metric spaces with continuous transformations T, which we will need for the proof of the ergodic theorem with Möbius weights. Thereafter, we prove the namely weighted ergodic theorem. We give a sufficient condition for Sarnak’s conjecture to hold for a given dynamical system, which we make use of in the following chapter. Thereupon, it is varified that Sarnak’s conjecture holds for the Thue–Morse shift and for skew product extensions of rational rotations on the circle. Lastly, it is shown that Sarnak’s conjecture from one of Chowla.

Pour construire des intervalles de confiance, nous pouvons utiliser diverses approches bootstrap. Nous avons un problème pour le contexte spécifique d’un paramètre de proportion lorsque l’estimateur usuel, la proportion de succès dans l’échantillon ˆp, est nul. Dans un contexte classique d’observations indépendantes et identiquement distribuées (i.i.d.) de la distribution Bernoulli, les échantillons bootstrap générés ne contiennent que des échecs avec probabilité 1 et les intervalles de confiance bootstrap deviennent dégénérés en un seul point, soit le point 0. En contexte de population finie, nous sommes confrontés aux mêmes problèmes lorsqu’on applique une méthode bootstrap à un échantillon de la population ne contenant que des échecs. Une solution possible s’inspire de l’estimateur utilisé dans les méthodes de [Wilson, 1927] et [Agresti et Coull, 1998] où ceux-ci considèrent ˜p l’estimateur qui prend la proportion de succès d’un échantillon augmenté auquel on a ajouté deux succès et deux échecs. La solution que nous introduisons consiste à effectuer le bootstrap de la distribution de ˆp mais en appliquant les méthodes bootstrap à l’échantillon augmenté de deux succès et deux échecs, tant en statistique classique que pour une population finie. Les résultats ont démontré qu’une version de la méthode percentile est la méthode bootstrap la plus efficace afin d’estimer par intervalle de confiance un paramètre de proportion autant dans un contexte i.i.d. que dans un contexte d’échantillonnage avec le plan aléatoire simple sans remise. Nos simulations ont également démontré que cette méthode percentile pouvait compétitionner avantageusement avec les meilleures méthodes traditionnelles. / A few bootstrap approaches exist to create confidence intervals. Some difficulties appear for the specific case of a proportion when the usual estimator, the proportion of success in a sample, is 0. In the classical case where the observations are independently and identically distributed (i.i.d.) from a Bernoulli distribution, the bootstrap samples only contain zeros with probability 1 and the resulting bootstrap confidence intervals are degenerate at the value 0. We are facing the same problem in the survey sampling case when we apply the bootstrap method to a sample with all observations equal to 0. A possible solution is suggested by the estimator found in the confidence intervals of [Wilson, 1927] and [Agresti et Coull, 1998] where they use ˜p the proportion of success in a augmented sample consisting of adding two successes and two failures to the original sample. The proposed solution is to use the bootstrap method on ˆp but where the bootstrap is based on the augmented sample with two additional successes and failures, whether the sample comes from i.i.d. Bernoulli variables or from a simple random sample. Results show that a version of the percentile method is the most efficient bootstrap method to construct confidence intervals for a proportion both in the classical setting or in the case of a simple random sample. Our results also show that this percentile interval can compete with the best traditional methods.