La compréhension automatique de la parole est au confluent des deux grands domaines que sont la reconnaissance automatique de la parole et l'apprentissage automatique. Un des problèmes majeurs dans ce domaine est l'obtention d'un corpus de données conséquent afin d'obtenir des modèles statistiques performants. Les corpus de parole pour entraîner des modèles de compréhension nécessitent une intervention humaine importante, notamment dans les tâches de transcription et d'annotation sémantique. Leur coût de production est élevé et c'est la raison pour laquelle ils sont disponibles en quantité limitée.Cette thèse vise principalement à réduire ce besoin d'intervention humaine de deux façons : d'une part en réduisant la quantité de corpus annoté nécessaire à l'obtention d'un modèle grâce à des techniques d'apprentissage semi-supervisé (Self-Training, Co-Training et Active-Learning) ; et d'autre part en tirant parti des réponses de l'utilisateur du système pour améliorer le modèle de compréhension.Ce dernier point touche à un second problème rencontré par les systèmes de compréhension automatique de la parole et adressé par cette thèse : le besoin d'adapter régulièrement leurs modèles aux variations de comportement des utilisateurs ou aux modifications de l'offre de services du système / Two wide research fields named Speech Recognition and Machine Learning meet with the Automatic Speech Language Understanding. One of the main problems in this domain is to obtain a sufficient corpus to train an efficient statistical model. Such speech corpora need a lot of human involvement to transcript and semantically annotate them. Their production cost is therefore quite high and they are difficultly available.This thesis mainly aims at reducing the need of human intervention in two ways: firstly, reducing the amount of corpus needed to build a model thanks to some semi-supervised learning methods (Self-Training, Co-Training and Active-Learning); And lastly, using the answers of the system end-user to improve the comprehension model.This last point addresses another problem related to automatic speech understanding systems: the need to adapt their models to the fluctuation of end-user habits or to the modification of the services list offered by the system

Numerical information is ubiquitous and people are continuously engaged in evaluating it by means of intuitive statistical judgments. Much research has evaluated if people’s judgments live up to the norms of statistical theory but directed far less attention to the cognitive processes that underlie the judgments. The present thesis outlines, compares, and tests two cognitive models for intuitive statistical judgments, summarized in the metaphors of the lazy and eager intuitive statistician. In short, the lazy statistician postpones judgments to the time of a query when the properties of a small sample of values retrieved from memory serve as proxies for population properties. In contrast, the eager statistician abstracts summary representations of population properties online from incoming data. Four empirical studies were conducted. Study I outlined the two models and investigated whether an eager or a lazy statistician best describes how people make intuitive statistical judgments. In general the results supported the notion that people spontaneously engage in a lazy process. Under certain specific conditions, however, participants were able to induce abstract representations of the experienced data. Study II and Study III extended the models to describe naive point estimates (Study II) and inference about a generating distribution (Study III). The results indicated that both the former and the latter type of judgment was better described by a lazy than an eager model. Finally, Study IV, building on the support in Studies I-III, investigated boundary conditions for a lazy model by exploring if statistical judgments are influenced by common memory effects (primacy and recency). The results indicated no such effects, suggesting that the sampling from long-term memory in a lazy process is not conditional on when the data is encountered. The present thesis makes two major contributions. First, the lazy and eager models are first attempts at outlining a process model that could possibly be applied for a large variety of statistical judgments. Second, because a lazy process imposes boundary conditions on the accuracy of statistical judgments, the results suggest that the limitations of a lazy intuitive statistician would need to be taken into consideration in a variety of situations.

This thesis addresses the challenges of large software and data-intensive systems. We will discuss a Big Data software that consists of quite a bit of Linux configuration, some Scala coding and a set of frameworks that work together to achieve the smooth performance of the system. Moreover, the thesis focuses on the Apache Spark framework and the challenging of measuring the lazy evaluation of the transformation operations of Spark. Investigating the challenges are essential for the performance engineers to increase their ability to study how the system behaves and take decisions in early design iteration. Thus, we made some experiments and measurements to achieve this goal. In addition to that, and after analyzing the result we could create a formula that will be useful for the engineers to predict the performance of the system in production.

Exact real arithmetic systems can specify any amount of precision on the output of the computations. They are used in a wide variety of applications when a high degree of precision is necessary. Some of these applications include: differential equation solvers, linear equation solvers, large scale mathematical models, and SMT solvers. This dissertation proposes a new exact real arithmetic system which uses lazy list of floating point numbers to represent the real numbers. It proposes algorithms for basic arithmetic computations on these structures and proves their correctness. This proposed system has the advantage of algorithms which can be supported by modern floating point hardware, while still being a lazy exact real arithmetic system.

In this thesis an analytical modelling approach is employed to predict and gain insight into the flows induced by turbulent plumes and jets above slender horizontal slots, in otherwise quiescent uniform environments. To supplement the solutions, the effect on the environment of a plume driven by an off-source supply of buoyancy was also considered. The solutions derived provide an advancement on existing idealised models for the jet and plume induced flows, and moreover, complement a number of key advances that have been made in our understanding of plume flows in recent years. The theory of functions of a complex variable, which has not previously been applied in such an application, has been utilised as a fundamental tool throughout the work. This has enabled the entrainment behaviour and geometry of the plumes to be accounted for when developing the induced flow solutions. A novel conformal mapping has been devised specifically to account for the curved perimeter of the contracting lazy plume. This modelling approach is robust in that future developments to aspects of the modelling, for instance, the formulation of a new entrainment closure, can be straightforwardly accounted for using the method. The induced flow solutions exhibit a range of flow patterns which are dependent on the source Richardson number of the plume flow. A pure plume induces a uniform horizontal flow. Forced and lazy plumes correspond to a relative deficit and excess in source buoyancy flux compared to the pure plume, respectively. Generally, forced plumes induce downwardly inclined flows, in contrast to lazy plumes, which induce upwardly inclined flows. Consistent with these solutions, the notionally lazy plume driven by a vertical uniform off-source supply of buoyancy induces an upwardly inclined flow. In addition to an improved understanding of induced flows, our solutions have provided us with insight into the plume flow. Notably, the solution corresponding to the forced plume has led us to fundamentally question existing models describing the plume and, in particular, closures that have been employed to model entrainment. We find that the existing well accepted closures exhibit some form of non-physical flow behaviour.

The answer to what Zaježová is, can be simple. Zaježová are so called „lazy“. Zaježová is perceived by the outside world - society - primarily as an alternative community of people living in coexistence with nature, and “Zaježová” presents itself “in this way”. Personally, I think that Zaježová is defined mainly by the strong individualities of people who are able to say about themselves and present that they live in a community. Last but not least, Zaježová is a place where I spent a large part of my childhood, because my parents belonged to these strong individuals and decided to live life “alone”. My work tries to capture what Zaježová is through various media, including architecture. The output is a set of atypical tourist maps and a proposal to transform the former fire station into a gallery.

The answer to what Zaježová is, can be simple. Zaježová are so called „lazy“. Zaježová is perceived by the outside world - society - primarily as an alternative community of people living in coexistence with nature, and “Zaježová” presents itself “in this way”. Personally, I think that Zaježová is defined mainly by the strong individualities of people who are able to say about themselves and present that they live in a community. Last but not least, Zaježová is a place where I spent a large part of my childhood, because my parents belonged to these strong individuals and decided to live life “alone”. My work tries to capture what Zaježová is through various media, including architecture. The output is a set of atypical tourist maps and a proposal to transform the former fire station into a gallery.

No description available.

No description available.

Yes / Case-based reasoning (CBR) is a nature inspired paradigm of machine learning capable to continuously learn from the past experience. Each newly solved problem and its corresponding solution is retained in its central knowledge repository called case-base. Withρ the regular use of the CBR system, the case-base cardinality keeps on growing. It results into performance bottleneck as the number of comparisons of each new problem with the existing problems also increases with the case-base growth. To address this performance bottleneck, different case-base maintenance (CBM) strategies are used so that the growth of the case-base is controlled without compromising on the utility of knowledge maintained in the case-base. This research work presents a hybrid case-base maintenance approach which equally utilizes the benefits of case addition as well as case deletion strategies to maintain the case-base in online and offline modes respectively. The proposed maintenance method has been evaluated using a simulated model of autonomic forest fire application and its performance has been compared with the existing approaches on a large case-base of the simulated case study. / Authors acknowledge the internal funding support received from Namal College Mianwali to complete the research work.