The objective of this work was the development of structured zeolite catalysts by growing of ZSM-5 and Y zeolites layers on the pre-treated FeCrAlloy wires, which could now offer technical advantage in catalytic application. The advantages of implementation of zeolitic coatings in industrial applications are that they have; lower pressure drop, high heat and mass transfer rates compared to standard pelleted or extruded catalysts. The key focus of this research was the generation of thin films of zeolite ZSM–5 and Y zeolite catalysts on the surface of a FeCrAlloy metal substrate. Using in-situ hydrothermal synthesis, the influence of the synthesis parameters such as substrate oxidation and crystallisation time on the zeolite crystallisation process in both the bulk phase (powder) and on the structured zeolite was studied and optimised. Then powder and structured Na-ZSM-5 and Na-Y were treated by calcination and ion exchange in post-synthesis treatment. Further post-synthesis modification was required in the zeolite Y case to improve the catalytic properties. The post synthetic modification of zeolite Y was carried out using acidified ammonium nitrate which was optimised to produce dealuminated zeolite Y with good crystallinity and a Si/Al = 8. Characterisation was performed after each stage of this work to optimise catalyst development using XRD, SEM, EDAX, BET, MAS-NMR, and TGA. Once the optimised zeolite Y and ZSM-5 structured catalysts prepared, cracking of n-heptane was carried out to assess the in catalytic performance compared with Y and ZSM-5 pellets in a fixed-bed reactor under the same operation conditions. The cracking of n–heptane over the pellets and structured catalysts for both ZSM–5 and Y zeolite showed very similar product selectivities for similar amounts of catalyst with apparent activation energy of around 60 kJ mol-1. This research demonstrates that structured catalysts can be manufactured with excellent zeolite adherence and when suitably activated/modified give comparable cracking results to the pelleted powder forms. These structured catalysts will improve temperature distribution in highly exothermic and endothermic catalysed processes.

An important stage in the policy process involves deciding what strategy is to be adopted for implementation so that the objectives of the policy are met in the best way possible. A Policy Implementation Strategy (PIS) adopts a broad view of implementation, which is argued to transcend formulation and decision-making, thereby offering a more realistic view of the policy process. Governmental decision-makers are often faced with having to choose one PIS amongst several possible alternatives, at varying cost levels. In order to aid in such a decision-making process, PIS effectiveness forecasts are proposed as a decision-support tool.Current methods for such a purpose are found to include ex-ante evaluative techniques such as Impact Assessment (IA) and Cost-Benefit Analysis (CBA). However, these approaches are often resource-intensive and such an investment is not always rewarded with accurate predictions. Hence, a judgmental forecasting approach for making PIS effectiveness predictions is proposed as a means for screening the different PIS under contention to provide a shortlist of candidates with particular potential. The selected few can then be further analysed via the quantitative evaluative techniques such as IA and CBA. Judgmental approaches to forecasting are considered ideal for such a role because they are relatively quick and inexpensive to implement. More specifically, a structured analogies approach is proposed as information about analogous PIS is believed to be useful for such a purpose.The proposed structured analogies approach is tested over a series of experiments and the evidence suggests that a structured analogies approach is more accurate when compared to unaided judgment and the more support given to the expert the better. Furthermore, experts were seen to produce considerably more accurate predictions than non-experts. Level of experience and number of analogies recalled did not seem to affect accuracy. The expert forecasts were also comparable to those produced by governments. The thesis concludes with suggestions for future research in the area.

This thesis aims to determine energy compensation following exercise induced energy expenditure (ExEE). The specific objectives were: I) to determine the impact of the time spent performing physical activity (PA) of varying intensities on body weight and composition (Study 1); II) to determine the overall energy compensation and the major predictors of energy compensation through the systematic review approach (Study 2); III) to develop new methods to measure energy intake (EI) (Study 3) and time spent performing different activities (Study 4); IV) to determine the effects of a lower (LI) and higher intensity (HI) ExEE intervention on energy compensation (Study 5); and V) to investigate the inter-individual variability regarding exercise induced energy compensation (Study 6). In Study 1, women spending more time performing light-intensity PA were shown to have lower adiposity compared to women spending more time performing moderate- and high-intensity PA. Results from Study 2 (systematic review) show an overall energy compensation of 25% following exercise interventions and that fat mass (FM), exercise intensity and duration of the intervention are the main predictors of energy compensation. To better capture energy compensation (i.e., EI and EE), new methods to measure EI and time spent performing activities were developed (Studies 3 and 4) and used in the following studies. In Study 5, overweight/obese women training at HI displayed higher energy compensation when compared to women training at LI, which was accompanied by a reduction of NSPA (non-structured physical activity) and a greater amount of time spent lying down. Results from Study 6 showed that complete compensators (CC) had higher EI, fat and carbohydrate intake at the onset of the ExEE intervention when compared to incomplete compensators (IC). However, the results also showed that dietary disinhibition was increased, whereas NSPA was decreased at the end of the intervention in IC. Taken together, these studies emphasize that weight loss following exercise is impeded by energy compensation. In addition to the impact of FM, exercise intensity and duration of the intervention on energy compensation, NSPA and cognitive factors also seem to modify energy compensation that occurs as a result of exercise.

Currently, the companies are exposed to many kinds of risk during the business, especially in case of activities which are overlapping the frontiers of inland market. Risk-aversive businessmen tend to minimize or even eliminate those risks so that they use different types of hedging instruments. The derivatives are a good choice but they are not able to meet all clients' needs sufficiently in their classic simple form at present. That is the reason for rise of their new combinations and modifications which are so-called "tailor-made" for each company separately. The purpose of this paper is to create the compact overview about those "second generation" products and show the interest and currency risk hedging options.

Hydrogen storage for stationary and mobile applications is an expanding research topic. Using liquid organic hydrides for hydrogen storage is one of the most promising alternatives as it provides simple and safe handling. Liquid organic hydrides are largely compatible with current transport infrastructure, whereas alternatives such as liquid and gaseous hydrogen and metal hydrides would require a completely new infrastructure. An attractive storage system is the so-called MTH system (Methylcyclohexane, Toluene and Hydrogen). The dehydrogenation of methylcyclohexane is a highly endothermic reaction. To improve the reaction kinetics, this research was to develop a structured catalyst with a conductive metal support (Fecralloy) which could hold an adherent catalytic washcoat (γ - Al2O3). The active phase was impregnated onto this support and the developed catalyst was tested for the dehydrogenation of methylcyclohexane. The catalyst preparation involved three key steps which were support oxidation, loading of an adherent washcoat and finally impregnation of the active phase. The oxidation and washcoat stages required significant optimisation. The optimum oxidation conditions were found to be 950 °C for 10 h. The washcoating procedure was optimised by modifying a one-step hybrid washcoating method suggested in patent literature. Characterization techniques including SEM, XRD and EDX were used to study each step of catalyst preparation. In addition the technique of STEM was used to study platinum dispersion on the catalytic washcoat. Finally the catalytic activity of the developed catalyst was compared with an in-house pelleted catalyst based on the material used to prepare the structured catalyst and commercially available platinum on γ - Al2O3. Three key factors: activity, selectivity and stability were evaluated. The activity and selectivity were studied at varied operating conditions of T = 340 °C - 400 °C, W/F = 7345 - 14690 g s/mol, H2/MCH molar ratio = 0 - 9 and P = 1.013 bar. The dehydrogenation reaction of methylcyclohexane was found to be very selective to toluene (above 99%). Compounds, which are considered coke precursors, were identified, to attempt to explain the mechanism of catalyst deactivation. By-product distribution was monitored and possible reaction pathways were postulated. To gauge the stability of the catalyst, long term life tests were also performed on the structured catalyst at 400 °C and W/F = 14690 g s/mol for approximately 400 h. The stability study investigated the different types of deactivation mechanisms. The catalyst evaluation study helped identify the effect of the alloy support, the alumina washcoat and platinum dispersion on the selectivity of the catalyst.

Cette thèse est organisée en deux parties indépendantes. La première partie s'intéresse à l'estimation convexe de matrice en prenant en compte à la fois la parcimonie et le rang. Dans le contexte de graphes avec une structure de communautés, on suppose souvent que la matrice d'adjacence sous-jacente est diagonale par blocs dans une base appropriée. Cependant, de tels graphes possèdent généralement une matrice d'adjacente qui est aussi parcimonieuse, ce qui suggère que combiner parcimonie et range puisse permettre de modéliser ce type d'objet de manière plus fine. Nous proposons et étudions ainsi une pénalité convexe pour promouvoir parcimonie et rang faible simultanément. Même si l'hypothèse de rang faible permet de diminuer le sur-apprentissage en diminuant la capacité d'un modèle matriciel, il peut être souhaitable lorsque suffisamment de données sont disponible de ne pas introduire une telle hypothèse. Nous étudions un exemple dans le contexte multiple kernel learning localisé, où nous proposons une famille de méthodes a vaste-marge convexes et accompagnées d'une analyse théorique. La deuxième partie de cette thèse s'intéresse à des problèmes de détection d'objets ou de signaux structurés. Dans un premier temps, nous considérons un problème de test statistique, pour des modèles où l'alternative correspond à des capteurs émettant des signaux corrélés. Contrairement à la littérature traditionnelle, nous considérons des procédures de test séquentielles, et nous établissons que de telles procédures permettent de détecter des corrélations significativement plus faible que les méthodes traditionnelles. Dans un second temps, nous considérons le problème de localiser des objets dans des images. En s'appuyant sur de récents résultats en apprentissage de représentation pour des problèmes similaires, nous intégrons des features de grande dimension issues de réseaux de neurones convolutionnels dans les modèles déformables traditionnellement utilisés pour ce type de problème. Nous démontrons expérimentalement que ce type d'approche permet de diminuer significativement le taux d'erreur de ces modèles. / This thesis is organized in two independent parts. The first part focused on convex matrix estimation problems, where both rank and sparsity are taken into account simultaneously. In the context of graphs with community structures, a common assumption is that the underlying adjacency matrices are block-diagonal in an appropriate basis. However, these types of graphs are usually far from complete, and their adjacency representations are thus also inherently sparse. This suggests that combining the sparse hypothesis and the low rank hypothesis may allow to more accurately model such objects. To this end, we propose and analyze a convex penalty to promote both low rank and high sparsity at the same time. Although the low rank hypothesis allows to reduce over-fitting by decreasing the modeling capacity of a matrix model, the opposite may be desirable when enough data is available. We study such an example in the context of localized multiple kernel learning, which extends multiple kernel learning by allowing each of the kernels to select different support vectors. In this framework, multiple kernel learning corresponds to a rank one estimator, while higher-rank estimators have been observed to increase generalization performance. We propose a novel family of large-margin methods for this problem that, unlike previous methods, are both convex and theoretically grounded. The second part of the thesis is about detection of objects or signals which exhibit combinatorial structures, and we present two such problems. First, we consider detection in the statistical hypothesis testing sense, in models where anomalous signals correspond to correlated values at different sensors. In most existing work, detection procedures are provided with a full sample of all the sensors. However, the experimenter may have the capacity to make targeted measurements in an on-line and adaptive manner, and we investigate such adaptive sensing procedures. Finally, we consider the task of identifying and localizing objects in images. This is an important problem in computer vision, where hand-crafted features are usually used. Following recent successes in learning ad-hoc representations for similar problems, we integrate the method of deformable part models with high-dimensional features from convolutional neural networks, and shows that this significantly decreases the error rates of existing part-based models.

Master of Science / Department of Computer Science / Doina Caragea / An Introduction to Databases course involves learning the concepts of data storage, manipulation, and retrieval. Relational databases provide an ideal learning path for understanding database concepts. The Structured Query Language (SQL) is a standard language for interacting with relational database. Each database vendor implements a variation of the SQL standard. Furthermore, a particular question that asks for some data can be written in many ways, using somewhat similar or structurally different SQL queries. Evaluation of SQL queries for correctness involves the verification of the SQL syntax and semantics, as well as verification of the output of queries and the usage of correct clauses. An evaluation tool should be independent of the specific database queried, and of the nature of the queries, and should allow multiple ways of providing input and retrieving the output. In this report, we have developed an evaluation tool for SQL queries, which checks for correctness of MySQL and PostgreSQL queries with the help of a parser that can identify SQL clauses. The tool developed will act as a portal for students to test and improve their queries, and finally to submit the queries for grading. The tool minimizes the manual effort required while grading, by taking advantage of the SQL parser to check queries for correctness, provide feedback, and allow submission.

This thesis explores boosting to improve decision-making in the context of the COVID-19 pandemic using a structured introspection. Structured introspection is an intervention where individuals are prompted with and are asked to estimate the importance of a set of attributes relevant to the decision in order to limit the prevalence of potential cognitive biases. To test the intervention, 281 participants divided into an intervention and control group answered an online survey with a dilemma about COVID-19. The dilemma was whether Sweden should shut down the economy or keep it open during the COVID-19 pandemic. The intervention group was asked to rate how important the attributes “saving lives”, “saving the economy”, “concern for the health of the elderly and risk groups”, and “concern for the quality of life and well-being of all citizens” should be for their decision. The control group was only prompted with the question and asked to think carefully. All participants were asked a set of control variables such as risk perception for self and others and emotions when thinking about COVID-19. The results did not show a significant influence on choice on decisions based on the intervention. They did however show a significant correlation with choice on risk perception as well as a correlation between choice on the dependent variable and the attributes in the intervention group.             The conclusion of the thesis is that structured introspection may not be suitable on a contemporary issue affecting participants directly, as they may already have strong opinions about the issue. Further and broader research needs to be conducted to determine in which circumstances this boost can be effective.

No description available.

Introduction: Innovations in modern medicine have led to the development of many effective laboratory tests for diagnosis and monitoring of disease. However, research suggests that 20-50% of tests may be inappropriately ordered. Test-ordering intervention development experts i.e. those staff at individual institutions who are tasked with designing different approaches to improving test-ordering practice, often have large datasets on the test-ordering practices within their jurisdiction as part of clinical care, but often have no clear guidance on how to choose which tests are worthy of intervention. These staff can include both clinicians and knowledge translation experts working in separately or in tandem to improve test-ordering behaviours at their institutions. The overall aim of this research is to build a prioritization framework to assist healthcare organizations in deciding which tests are worth targeting for intervention. Methods: The first stage of this project was a scoping review of the literature, the purpose of which was to determine the factors and processes used by intervention developers to choose which tests to target for intervention. Identification of these articles allowed for the descriptive analysis of the factors and processes reported in making the test-ordering decisions. The second stage involved semi-structured interviews with intervention development experts and contains information on their experiences with measuring and prioritizing among factors, as well as which processes they found to be most effective in making the decision on which tests to target for intervention. Results: By exploring all relevant guideline statements and related empirical studies, we gathered a wide range of factors to consider when deciding which test(s) to target for intervention. Overall, we identified 18 factors in our review and 30 in our interviews, highlighting the potential complexity of these decisions. While our studies were not the first to demonstrate that targeting tests is complex and that many factors must go into the decision of deciding which tests to target, our work has identified the most comprehensive range of factors available. We also explored processes reported by the studies that led to these decisions. Of the nine processes identified in our review, some were identified in a majority of cases (literature review, followed by clinical standard-implicit and consensus process), and other processes far more infrequently. Our interviews with relevant stakeholders demonstrated that the interviewees used 18 unique processes to decide which tests to target for intervention and helped to prioritize among the processes that are most effective in making these decisions. Conclusion: The current work prepares the way to develop a framework designed to help intervention developers choose which tests can most efficiently result in improved test-ordering processes. After additional interviews to ensure saturation of themes, we will be able to proceed with framework development, perhaps involving a consensus process of all relevant stakeholders. We hope to widely distribute our framework to assist intervention development experts working in a wide variety of milieus to help them decide which tests are worth targeting for intervention such that their respective institutions can provide the highest quality of care to patients.