This study analyzes the determination of the exchange rate system in a small economy when external real and monetary disturbances occur. Choice of exchange rate policy is investigated using a model assuming rational expectations and a loss function expressing the squared deviations of the small country output from desired output. The distinguishing feature of the analysis is the emphasis on real as well as monetary disturbances which originate abroad but are a source of domestic output variation. the link between foreign monetary and real disturbances and variance in output is traced using the thoretical model and the loss function assumed. The emphasis of the analysis is on a three country (one small and two large) trading situation, whereby the small country trades with two major large country trading partners. It is assumed throughout that there is perfect commodity arbitrage between two large countries. The small country imports an intermediate good from one of the large countries and exports a finished good. The small country doesnot import goods for consumption. there is perfect capital movement between two large countries, but capital is immobile between the small and these two large countries. The analysis indicates that occurrence of purely nominal shocks abroad are not transmitted to the small country under floating exchange rate system. The presence of real disturbances in large countries induce lower prices for the goods they produce, but the effect on the exchange rate is ambiguos. This study concludes that in general the adoption of a flexible exchange rate system by a small country is preferred and results in lower loss in most cases of external disturbance.

With the increase in the occurrence of disasters (natural and man-made) that leave people injured, handicapped or dead, the disaster management theory is gaining more importance. As a consequence, human assistance and disaster relief organizations are managing increasingly more inventories anticipated to help people in need. Donations are the common means used by humanitarian relief organizations for procuring commodities to support some of their programs. Previous experiences have indicated that donations become a burden instead of offering relief when they do not match actual victims' needs. Accepting or rejecting donations is a key issue that can produce not only economic losses but loss of lives as well. The objective of this thesis is to provide a means of assessing acceptance or rejection decisions using decision tree analysis theory and utility theory. The proposed model considers the inputs that a decision-maker may face when accepting or rejecting a donation. Such inputs include these categories: the probability of the occurrence of disaster, the need for and further use of a commodity, the unit price and holding cost of the item, the benefit provided by the donation, and the probability of having subsequent donations when the initial donation is initially rejected. Various scenarios are simulated in Excel® environment through the Monte Carlo process. This will assess the varied impacts from the alternative inputs in the decision making process; a sensitivity analysis will evaluate the effects of various decisions. The results obtained from the simulation of the diverse scenarios indicate that the decision of accepting or rejecting donations is driven more by the possibility of the use of the commodity than by the probability of occurrence of the disaster. The findings from the model also indicate that the decision of accepting or rejecting is more sensitive to the relationship of sale price to benefit deployment of the commodity than to sale price alone. The simulation of the expected monetary benefit of the relief provided results in the development of graphs that can affect the decision making process when accepting or rejecting donations.

No description available.

The thesis reviews the literature relating to girls and computing within a framework which is structured around three specific questions. First, are there differences between girls and boys in their participation in class computing activities and/or in non-class computing activities? Second, do these differences in participation in computing activities have broader implications which justify the growing concern about the under-representation of girls? Third, wahy are girls under-represented in these activities? Although the available literature is predominantly descriptive, the underlying implicit theoretical model is essentially a social learning model. Girl's differential participation is attributed to learned attitudes towards computing rathan to differences between girls and boys in general ability. These attitudes, which stress the masculine, mathematical, technological aspects of computing are developed through modelling, direct experience, intrinsic and extrinsic reinforcement and generalisation from pre-existing, attitudes to related curriculum areas. In the literature it is implicitly assumed that these attitudes underlie girl's decisions to self-select out of computing activities. In this thesis predictions from a social learning model are complemented by predictions derived from expectancy-value, cognitive dissonance and self-perception theories. These are tested in three separate studies. Study one provides data from a pretest-posttest study of 24 children in a year four class learning BASIC. It examines pre- and posttest differences between girls and boys in computing experience, knowledge and achievement as well as the factors relating to computing achievement. Study two uses a pretest-posttest control group design to study the gender differences in the impact of the introduction of Logo into years 1, 3, 5 and 7 in both a coeducational and single-sex setting using a sample of 222 children from three schools. Study three utilises a larger sample of 1176 students, drawn from three secondary schools and five primary schools, enabling an evaluation of gender differences in relation to a wide range of class computing experiences and in a broader range of school contexts. The overall results are consistent across the three studies, supporting the contention that social factors, rather than ability differences influence girls' participation and achievement in computing. The more global theoretical framework, drawing on social learning, expectancy-value, cognitive dissonance and self-perception theories, provides a more adequate explanation of gender differences in participation than does any one of these models.

<p>Generating and stabilizing periodic motions in nonlinear systems is a challenging task. In the control system community this topic is also known as limit cycle control. In recent years a framework known as Virtual Holonomic Constraints (VHC) has been developed as one of the solutions to this problem. The aim of this thesis is to give an insight into this approach and its practical application.</p><p>The contribution of this work is primarily the experimental validation of the theory. A step by step procedure of this methodology is given for motion planning, as well as for controller design. Three particular setups were chosen for experiments: the inertia wheel pendulum, the Furuta pendulum and the two-link planar pendulum. These under-actuated mechanical systems are well known benchmarking setups for testing advanced control design methods.</p><p>Further application is intended for cases such as biped robot walking/running, human and animal locomotion analysis, etc.</p>

<p>Den svenska polisens arbete är kunskapsintensivt. Organisationen hanterar dagligen stora mängder information. Särskilt ordningspoliser har ett stort informationsbehov för att kunna agera på lämpligt sätt i olika situationer. Handlandet bygger på kännedom om lagrum och föreskrifter, men också uppgifter om medborgarna och vetskap om kollegors aktiviteter. Denna kunskap kan i stor utsträckning fås genom de informationssystem som finns inom organisationen. Ordningspoliser har dock dålig tillgång till dem i fält vilket gör att den viktiga koordinationen och dokumentationen av arbetet inte är så effektiv som den skulle kunna vara.</p><p>Denna uppsats undersöker vilka informationsbehov ordningspoliser har i fält samt hur ett mobilt IT-stöd skulle kunna se ut för att stödja detta behov. Vi intresserar oss för ordningspolisens arbetssätt och för hur användandet av en mobil applikation skulle kunna gå till. Vi diskuterar vilka funktioner och resurser som skulle vara lämpliga att ha tillgång till i olika situationer och vilken nytta och återverkningar dessa skulle ge på arbetssättet.</p><p>Vi har funnit att ordningspolisen har ett stort och varierande informationsbehov, samtidigt som spridningen av information är begränsad i verksamheten. Ett mobilt IT-stöd skulle kunna öka polismannens kunskap och arbetsmotivation. Det finns ett antal informationsresurser som är intressanta att ha tillgång till. Användandet påverkas av hur den yttre miljön ser ut samt av arbetsuppgiftens karaktär. Särskilt det patrullerande arbetet skulle gagnas av ett mobilt IT-stöd och ge patrullerna större möjlighet att agera självständigt. Även arbete på en brottsplats skulle kunna effektiviseras genom att en polisman exempelvis får stöd och tips på lämpliga lagrum och möjliga åtgärder.</p>

<p>After decades of power generating units increasing in size, there is currently a growing focus on distributed generation, power generation close to energy loads. Investments in large-scale units have been driven by economy of scale, but recent technological improvements on small generating plants have made it possible to exploit the benefits of local power generation to a larger extent than previously. Distributed generation can improve power system efficiency because heat can be recovered from thermal units to supply heat and thermally activated cooling, and because small-scale renewables have a promising end-user market. Further benefits of distributed generation include improved reliability, deferral of often controversial and costly grid investments and reduction of grid losses. The new appeal of small-scale power generation means that there is a need for new tools to analyze distributed generation, both from a system perspective and from the perspective of potential developers. In this thesis, the focus is on the value of power generation for end-users. The thesis identifies how an end-user can find optimal distributed generation systems and investment strategies under a variety of economic and regulatory scenarios. The final part of the thesis extends the analysis with a bottom-up model of how the economics of distributed generation for a representative set of building types can transfer to technology diffusion in a market.</p><p>Four separate research papers make up the thesis. In the first paper, Optimal Investment Strategies in Decentralized Renewable Power Generation under Uncertainty, a method for evaluation of investments in renewable power units under price uncertainty is presented. It is assumed the developer has a building with an electricity load and a renewable power resource. The case study compares a set of wind power systems with different capacity and finds that capacity depends on the electricity price and that there under uncertain prices can be a significant value in postponing investment until larger projects are profitable. In the second paper, Combined Heat and Power in Commercial Buildings: Investment and Risk Analysis, a Monte Carlo simulation program to find the value and risk characteristics of combined heat and power units is presented. Using historical price data to estimate price process parameters, it is shown that uncertain prices should not be a barrier for investment, since on-site generators can adapt to uncertain prices and reduce the total energy cost risks. In, Optimizing Distributed Generation Systems for Commercial Buildings, which uses a mixed integer linear program, distributed generation portfolios that maximize profitability are tailored to a building's energy load. Distributed generation with heat recovery and thermally activated cooling are found profitable in an office and a health care building, using current generator data and energy tariffs from California. With the fourth paper, Distributed Energy Resources Market Diffusion Model, the analysis is taken a step further to predict distributed generation market diffusion. Market penetration is assumed to depend on economic attractiveness and knowledge and trust in the technologies. A case study based on the U.S. commercial sector depicts a large market for reciprocating engines and microturbines, with the West and Northeast regions driving market diffusion. Technology research and outreach programs can speed up and change the path of capacity expansion.</p><p>The thesis presents three different models for analyzing investments in distributed generation, all of which have benefits and disadvantages. Choice of model depends on the specific application, but the different approaches can be used on the same problem to analyze it from different viewpoints. The cases in the thesis indicate that distributed generation can reduce expected energy costs while at the same time improve cost predictability. Further, the thesis identifies several important factors and potential barriers to distributed generation adoption. Analyzing distributed generation from the end-user perspective is important also for policy makers, because of the importance of estimating how the market will react to potential policy measures. The thesis shows that small-scale generating capacity has the potential to increase in the near future. Further research should increase the understanding of economic and environmental issues related to distributed generation, while policy makers should aim to construct and implement measures that make it attractive for end-users to invest in efficient local generating capacity.</p>

<p>Syftet med uppsatsen är att undersöka om det förkommer underprissättning och om den består de första fem månaderna samt finna sannolikheten för en positiv utveckling efter börsintroduktionen. Eftersom vi har studerat kursförändringar efter börsintroduktioner har vi valt att använda en kvantitativ metodansats i form av en eventstudie. Teorierna behandlar främst förklaringar till varför en underprissättning sker samt hur publicitet och informations asymmetri påverkar prissättningen. Teorier som berörs är bland annat ”Winners Curse”, ”Cascade theory” och ”Signalling theory”.</p><p>Våra empiriska data kommer främst från bolagens prospekt samt kursinformation från Affärsvärlden och Stockholmsbörsen. Prospekten har visat kursintervall och introduktionsdatum, val av garant med mera. Affärsvärlden visar branschtillhörighet samt bransch- och indexhistorik. Stockholmsbörsen tillför slutkurser för samtliga bolag samt information om noteringskrav och -process.</p><p>Studien visar att de nyintroducerade bolagen har varit underprissatta med i genomsnitt 2,17% under perioden 2001-2006, vilket är betydligt lägre än vad tidigare studier visat. Detta beror främst på att de tidigare studierna har baserats på den amerikanska marknaden och under en tidsperiod då större underprissättning har förekommit än under 2001-2006. Den genomsnittliga utvecklingen efter 5 dagar, 1 månad, 3 månader, 5 månader respektive 1 år har varit 2,40%, 2,46%, 3,06%, 3,73% respektive 5,01%. Det har även förkommit stora skillnader mellan olika branscher, garanter, introduktionsår, introduktionspris och överteckningsgrad.</p> / <p>The purpose of this essay is to examine if it occurs under pricing and if it last the first five month and find the probability of a positive share development after a initial public offering. Since we have studied exchange fluctuation after initial public offerings have we chosen to use an quantitative method approach. The theories consider explanations to why under pricing occurs and how publicity and information asymmetry affect the price determination. ”Winners Curse”, ”Cascade theory” and ”Signalling theory” are some of the theories that are mentioned.</p><p>Our empirical data is mainly captured from the companies prospects and share price information from Affärsvärlden and the Stockholm Exchange. The prospects has shown share price intervals, initial public offering dates, underwriter etc. Affärsvärlden show type of business and index history. The Stockholm Exchange contribute closing prices for all of the studied companies and information about demands for quotation and the process regarding initial public offerings.</p><p>Our study show that initial public offerings have been under priced with an average of 2,17% during the period 2001-2006, which is considerably lower than earlier studies. The reason is mainly because the earlier studies has based on the American stock market and under a period when larger under pricing have existed. The average development after 5 days, 1 month, 3 months, 5 months respectively 1 year has been 2,40%, 2,46%, 3,06%, 3,73% respectively 5,01%. It has also occurred big differences between different type of business, underwriters, year of introduction, oversubscribtion and initial share price.</p>

The decentralized Markov decision process (Dec-POMDP) is a powerful formal model for studying multiagent problems where cooperative, coordinated action is optimal, but each agent acts based on local data alone. Unfortunately, it is known that Dec-POMDPs are fundamentally intractable: they are NEXP-complete in the worst case, and have been empirically observed to be beyond feasible optimal solution.To get around these obstacles, researchers have focused on special classes of the general Dec-POMDP problem, restricting the degree to which agent actions can interact with one another. In some cases, it has been proven that these sorts of structured forms of interaction can in fact reduce worst-case complexity. Where formal proofs have been lacking, empirical observations suggest that this may also be true for other cases, although less is known precisely.This thesis unifies a range of this existing work, extending analysis to establish novel complexity results for some popular restricted-interaction models. We also establish some new results concerning cases for which reduced complexity has been proven, showing correspondences between basic structural features and the potential for dimensionality reduction when employing mathematical programming techniques.As our new complexity results establish that worst-case intractability is more widespread than previously known, we look to new ways of analyzing the potential average-case difficulty of Dec-POMDP instances. As this would be extremely difficult using the tools of traditional complexity theory, we take a more empirical approach. In so doing, we identify new analytical measures that apply to all Dec-POMDPs, whatever their structure. These measures allow us to identify problems that are potentially easier to solve on average, and validate this claim empirically. As we show, the performance of well-known optimal dynamic programming methods correlates with our new measure of difficulty. Finally, we explore the approximate case, showing that our measure works well as a predictor of difficulty there, too, and provides a means of setting algorithm parameters to achieve far more efficient performance.

This thesis investigates the following question: Can supervised learning techniques be successfully used for finding better solutions to multistage stochastic programs? A similar question had already been posed in the context of reinforcement learning, and had led to algorithmic and conceptual advances in the field of approximate value function methods over the years. This thesis identifies several ways to exploit the combination "multistage stochastic programming/supervised learning" for sequential decision making under uncertainty. Multistage stochastic programming is essentially the extension of stochastic programming to several recourse stages. After an introduction to multistage stochastic programming and a summary of existing approximation approaches based on scenario trees, this thesis mainly focusses on the use of supervised learning for building decision policies from scenario-tree approximations. Two ways of exploiting learned policies in the context of the practical issues posed by the multistage stochastic programming framework are explored: the fast evaluation of performance guarantees for a given approximation, and the selection of good scenario trees. The computational efficiency of the approach allows novel investigations relative to the construction of scenario trees, from which novel insights, solution approaches and algorithms are derived. For instance, we generate and select scenario trees with random branching structures for problems over large planning horizons. Our experiments on the empirical performances of learned policies, compared to golden-standard policies, suggest that the combination of stochastic programming and machine learning techniques could also constitute a method per se for sequential decision making under uncertainty, inasmuch as learned policies are simple to use, and come with performance guarantees that can actually be quite good. Finally, limitations of approaches that build an explicit model to represent an optimal solution mapping are studied in a simple parametric programming setting, and various insights regarding this issue are obtained.