Analisam-se as questões envolvendo a modulação temporal da jurisprudência, como técnica adequada à solução do problema que orbita em torno das drásticas mudanças dos precedentes dos tribunais. Compara o sistema da common Law e da civil Law e a crescente tendência de absorção de institutos destes dois sistemas. Analisam-se exemplos de modulação de decisões no Direito Brasileiro, além das técnicas de modulação existentes no Direito Estrangeiro. Critica o modo como os tribunais brasileiros têm tratado a uniformização da jurisprudência. Observa critérios existentes no ordenamento jurídico brasileiro e que se prestam à aplicação como norte à modulação temporal, principalmente diante dos princípios da boa-fé, proteção a confiança e segurança jurídica. / It analyses the topics envolving juriprudence modulation as a technique adequate to solute the problem of the drastics overrulings. It compares both common law and civil law systems and the actual tendency to absorve institutes from both systems. It also analyses examples of jurisprudence modulation in brazilian law and techniques from other countries. It criticizes the way brazilian courts treat the jurisprudence uniformization. It observes the many criterions in brazilian law that allow to be applied as a direction to the jurisprudence modulation, specially in the matter of good faith, confidence protection and juridical security principles.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第22232号 / 理博第4546号 / 新制||理||1653(附属図書館) / 京都大学大学院理学研究科数学・数理解析専攻 / (主査)教授 望月 新一, 教授 玉川 安騎男, 准教授 星 裕一郎 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

<p>U disertaciji se navode definicije i svojstva malotalasne transformacije i kvaziasimptotskog<br />ponaˇsanja distribucija. Teoreme Abelovog i Tauberovog tipa<br />su koriˇs&acute;cene za asimptotsku analizu temperiranih distribucija, u odnosu na<br />njihovu malotalasnu transformaciju. Takode, prouˇcavana je i malotalasna<br />transformacija ultradiferencijabilnih funkcija i temperiranih ultradistribucija.</p> / <p> In thesis we give basic notions of wavelet transform and quasiasymptotic<br /> behavior of distributions. Via Abelian and Tauberian type of theorems we<br /> study quasiasymptotic behavior of tempered distributions, related to their<br /> wavelet transform. Further, we study wavelet transform of ultradifferential<br /> functions and tempered ultradistributions.</p>

Sett över hela landet växer ädellövskog endast på en procent av den produktiva skogsmarken. I det studerade området Dalsland utgörs virkesförrådet uppskattningsvis av tre procent ädla lövträd. Dess förekomst är därför att betrakta som relativt sällsynt. Denna studie hade för avsikt att intervjua markägare som har valt att nyetablera med ädellövskog och genom det förstå vad det är som bidragit till att de föryngrat med dessa trädslag. Undersökningen utfördes genom att sex individer som nyetablerat med ädellövskog intervjuades. Resultatet sammanställdes därefter genom att se på gruppen som en helhet och lyfta fram samtliga påverkande faktorer. Utifrån resultatet kunde 13 faktorer identifieras som bidragande och en förståelse kring dem redogöras för.  Det kan konstateras att ett flertal faktorer påverkat markägarna i valet att plantera ädellövskog.

The martensite structure is the most important microstructure in tool steel due to its high hardness. However, a lack of ductility is the major drawback. In order to improve the ductility and still maintaining a suitable hardness a tempering process is needed. The tempering process will cause recovery and recrystallization in the matrix, and moreover carbides will precipitate. The specific carbides have different characteristics and thus the type of carbide formed during tempering is very important for the properties of the steel. The simulation software (TC-Prisma) is interesting because it can predict type, size, and amount of carbides. The present study was carried out to investigate both the microstructure, hardness evolution of martensite and precipitation which occurred in Fe-C-Cr steel with different compositions, tempered at 700oC. The experimental results were compared with simulation results. Micro-Vickers hardness test with a load of 100 g was used and the hardness value dropped 40% and 60% in low carbon alloy and high carbon alloy steels, respectively. The significant drop occurred during the first 30 seconds of tempering due to recovery of the matrix. Hardness values slightly decreased and then stabilized during continued tempering. The microstructure of martensite and the morphology of carbides at different tempering times were examined by scanning and transmission electron microscopy in order to study the precipitation of carbides from the nucleation and growth to coarsening. There are three types of carbides which precipitated in the Fe-C-Cr specimens: M7C3, cementite and M3C2 depending on the composition. Fe-0.16C-4.05Cr contained M7C3, Fe-0.95C-1.065Cr contained cementite and M3C2 and Fe-014C-0.983Cr and Fe-0.88C-4.12Cr contained M7C3 and cementite. M7C3 has a faceted shape and precipitates referentially at grain boundaries. On the other hand, cementite has an elongated shape and precipitated mainly at grain boundaries but also intragranulary. M3C2 has a rounded shape and was seen only in very small amounts, and seemed to precipitate at random sites. The trend of carbide growth in experiments is consistent with the simulations using TC-Prisma, but more work is needed to enable quantitative comparisons.

No description available.

No description available.

No description available.

No description available.

Les tâches de vision artificielle telles que la reconnaissance d’objets demeurent irrésolues à ce jour. Les algorithmes d’apprentissage tels que les Réseaux de Neurones Artificiels (RNA), représentent une approche prometteuse permettant d’apprendre des caractéristiques utiles pour ces tâches. Ce processus d’optimisation est néanmoins difficile. Les réseaux profonds à base de Machine de Boltzmann Restreintes (RBM) ont récemment été proposés afin de guider l’extraction de représentations intermédiaires, grâce à un algorithme d’apprentissage non-supervisé. Ce mémoire présente, par l’entremise de trois articles, des contributions à ce domaine de recherche. Le premier article traite de la RBM convolutionelle. L’usage de champs réceptifs locaux ainsi que le regroupement d’unités cachées en couches partageant les même paramètres, réduit considérablement le nombre de paramètres à apprendre et engendre des détecteurs de caractéristiques locaux et équivariant aux translations. Ceci mène à des modèles ayant une meilleure vraisemblance, comparativement aux RBMs entraînées sur des segments d’images. Le deuxième article est motivé par des découvertes récentes en neurosciences. Il analyse l’impact d’unités quadratiques sur des tâches de classification visuelles, ainsi que celui d’une nouvelle fonction d’activation. Nous observons que les RNAs à base d’unités quadratiques utilisant la fonction softsign, donnent de meilleures performances de généralisation. Le dernière article quand à lui, offre une vision critique des algorithmes populaires d’entraînement de RBMs. Nous montrons que l’algorithme de Divergence Contrastive (CD) et la CD Persistente ne sont pas robustes : tous deux nécessitent une surface d’énergie relativement plate afin que leur chaîne négative puisse mixer. La PCD à "poids rapides" contourne ce problème en perturbant légèrement le modèle, cependant, ceci génère des échantillons bruités. L’usage de chaînes tempérées dans la phase négative est une façon robuste d’adresser ces problèmes et mène à de meilleurs modèles génératifs. / High-level vision tasks such as generic object recognition remain out of reach for modern Artificial Intelligence systems. A promising approach involves learning algorithms, such as the Arficial Neural Network (ANN), which automatically learn to extract useful features for the task at hand. For ANNs, this represents a difficult optimization problem however. Deep Belief Networks have thus been proposed as a way to guide the discovery of intermediate representations, through a greedy unsupervised training of stacked Restricted Boltzmann Machines (RBM). The articles presented here-in represent contributions to this field of research. The first article introduces the convolutional RBM. By mimicking local receptive fields and tying the parameters of hidden units within the same feature map, we considerably reduce the number of parameters to learn and enforce local, shift-equivariant feature detectors. This translates to better likelihood scores, compared to RBMs trained on small image patches. In the second article, recent discoveries in neuroscience motivate an investigation into the impact of higher-order units on visual classification, along with the evaluation of a novel activation function. We show that ANNs with quadratic units using the softsign activation function offer better generalization error across several tasks. Finally, the third article gives a critical look at recently proposed RBM training algorithms. We show that Contrastive Divergence (CD) and Persistent CD are brittle in that they require the energy landscape to be smooth in order for their negative chain to mix well. PCD with fast-weights addresses the issue by performing small model perturbations, but may result in spurious samples. We propose using simulated tempering to draw negative samples. This leads to better generative models and increased robustness to various hyperparameters.