No description available.

The forecasting of Eskom transmission electrical network demands is a complex task. The lack of historical data on some of the network components complicates this task even further. In this dissertation a model is suggested which will address all the requirements of the transmission system expansion engineers in terms of future loads and market trends. Suggestions are made with respect to ways of overcoming the lack of historical data, especially on the point loads, which is a key factor in modelling the electrical networks. A brief overview of the transmission electrical network layout is included to provide a better understanding of what is required from such a forecast. Lastly, some theory on multiple regression, neural networks and qualitative forecasting techniques is included, which will be of value for further model developments. / Computing / M. Sc. (Operations Research)

Die Modellkomponenten für parametrisierteWolkenphysik, Strahlung und Sichtweitenbestimmung im Nebelvorhersagemodell PAFOG, das kürzlich in Zusammenarbeit mit dem Deutschen Wetterdienst als lokales Vorhersagesystem entwickelt wurde und für die Kurzfristprognose eingesetzt werden kann, werden vorgestellt. Die Modellphilosophie orientiert sich an einer mathematisch-physikalisch fundierten Beschreibung der beteiligten meteorologischen Prozesse, deren Einzelheiten in dieser Arbeit diskutiert werden. / This paper presents the model components for parameterized cloud physics, radiation and visibility determination as implemented in the local forecast model PAFOG. PAFOG has been recently developed in cooperation with the GermanWeather Service DWD. PAFOG can be employed for short-range forecasts of radiation fog and visibility. The philosophy of the model strongly emphasizes a mathematically and physically based formulation of the involved meteorological processes the details of which are discussed in this paper.

Die im Nebelvorhersagemodell PAFOG enthaltenen Modellkomponenten für parametrisierte Wolkenphysik, Strahlung und Sichtweitenbestimmung wurden durch Module zur Beschreibung der Interaktion mit dem Boden und der Vegetation ergänzt. Das auf diese Weise komplettierte Modellsystem PAFOG-V kann dazu verwendet werden, das lokale Auftreten von Strahlungsnebel und niedriger stratiformer Bewölkung vorherzusagen. / The paper presents an extension of the model components for parameterized cloud physics, radiation and visibility determination as implemented in the local forecast model PAFOG to include the interaction with the soil and the vegetation. The resulting forecast system PAFOG-V can be used to predict local events of radiation fogs and of low level stratiform clouds.

Forecast of the pre-school education needs in Astana until 2030 Abstract This thesis aims to analyze future preschool education needs in Astana city. The low coverage of children by preschool education (a small number of preschool education institutions) initially lays the foundation of inequality elements and contradicts with the basic principles of the State education policy aimed at accessibility of education for the citizens of the country. As a consequence of insufficient governmental financing, the cost of children maintenance increased and it became the reason of unavailability of kindergarten service for separate groups of people. Consequently, the future preschool needs in Astana until 2030 is of particular interest. The thesis addresses issue of future population development dynamics in new capital city. This aim achieved with the help of forecasting techniques applied for Astana population until 2030. The thesis aimed to present the picture of how the trends (mortality, fertility and migration) are developing and how they will affect on the preschool needs in the future. Keywords: Astana, Kazakhstan, preschool needs, probability of dying, contributions to changes in life expectancy at birth, age specific fertility rate, gross reproduction rate, net reproduction rate, forecasting, cohort-component...

Sugere-se neste trabalho que o irrealismo constitui uma característica marcante da corrente dominante na ciência econômica que explica o desempenho decepcionante das atividades de previsão e explanação realizadas com base nesse paradigma. Após desenvolver o pouco claro conceito de irrealismo na ciência econômica, mostrando quando ele ocorre e quais são suas causas, busca-se relacionar sua presença com aquele desempenho decepcionante. Defende-se que ele pode ser explicado pela negligência da maior parte da corrente dominante em relação ao realismo de suas proposições, negligência traduzida na utilização amplamente difundida nas práticas dessa corrente de modelos irrealistas, que são aqueles que não buscam ou, se o fazem, não são bem-sucedidos em capturar uma parcela relevante da realidade. Sugere-se que o emprego de modelos irrealistas é geralmente - mas não sempre - o resultado da insistência do mainstream na ciência econômica em aderir à abordagem dedutivista em um mundo caracterizado pela não-ubiqüidade de regularidades estritas que ela invariavelmente pressupõe, e se manifesta tipicamente na aplicação generalizada em situações concretas de modelos econômicos fortemente abstratos cujos pressupostos implicam a operação de mecanismos que são inválidos nessas situações específicas. A explicação para esta tendência ao irrealismo do mainstream, por sua vez, se encontra no fato de que a grande maioria dos modelos elaborados a partir dessa perspectiva pressupõe a onipresença de estruturas de mercado competitivas a despeito de que em muitas - e mesmo na maioria das - situações reais elas estão claramente ausentes, o que decorre do viés ideológico que os economistas associados à corrente dominante possuem e que se caracteriza pela crença inequivoca nas insuperáveis virtudes do mecanismo de mercado e da propriedade privada como princípios orientadores centrais da organização da produção e mesmo da vida em sociedade. Essa crença se encontra na origem do liberalismo econômico tradicional e do neoliberalismo, e sua defesa obrigatoriamente requer que os mercados sejam, pelo menos em sua grande maioria, competitivos. Para satisfazer essa hipótese a ideologia neoliberal impõe à realidade a onipresença de estruturas de mercado competitivas, possíveis em abstrato mas geralmente ausentes em situações reais, o que faz com que os modelos construídos a partir da abordagem que a ela adere assim como a própria abordagem sejam freqüentemente irrealistas. / This work suggests that irrealism is a remarkable feature of mainstream in economics and explains the poor performance of both activities of prediction and explanation that heavily draw from this paradigm. After elaborating the unclear concept of irrealism in economics, showing when it occurs and what are its causes, I try to relate its presence to that poor performance. It\"s sustained that this performance can be explained by the negligence of most of mainstream practioners concerning the realism of their propositions, negligence that translates into the widely spread use of unrealistic models, those that do not try or, if they do, they don\"t succeed in capturing a relevant portion of the reality, in their practices. It\"s suggested that the use of such an unrealistic models is mostly - but not always - due to the insistence of mainstream economics in sticking to the deductivistic approach in a world where the strict regularities that it inevitably assumes are extremely scarce, and tipically shows itself in abstract models widely applied to concrete situations where their assumptions imply the operation of mechanisms that happen to be invalid in those specific situations. The account of that tendency to irrealism in mainstream economics should be looked for in the fact that the large majority of models they create assumes the ubiquity of competitive market structures despite in many real situations - probably in most of them - they are clearly absent, what is a result of the ideological bias that mainstream economists have, defined by the strong belief in the insurmountable virtues of the market mechanism and private property as general principles for the organization of production and even for life in society. This belief is found in the origin of traditional economical liberalism and of neoliberalism, and its defense inevitably requires markets to be, at least in their large majority, competitive. In order to satisfy that assumption neoliberal ideology imposes into reality the ubiquity of competitive market structures, possible as an abstraction but generally absent in real situations, what frequently making those models that heavily draw from this paradigm as well as the whole approach totally unrealistic.

This thesis comprises three main chapters focusing on a number of issues related to forecasting economic and nancial time series. Chapter 2 contains a detailed empirical study comparing forecast perfor- mance of a number of popular term structure models in predicting the UK yield curve. Several questions are addressed and investigated, such as whether macroeconomic information helps in forecasting yields and whether predict- ing performance of models change over time. We nd evidence of signi cant time-variation in forecast accuracy of competing models, particularly during the recent nancial crisis period. Chapter 3 explores density forecasts of the yield curve which, unlike the point forecasts, provide a full account of possible uncertainties surrounding the forecasts. We contribute by evaluating predictive performance of the recently developed stochastic-volatility arbitrage-free Nelson-Siegel models of Chris- tensen et al. (2010). The one-month-ahead predictive densities of the models appear to be inferior compared to those of their constant-volatility counter- parts. The advantage of modelling time-varying volatilities becomes evident only when forecasting interest rates at longer horizons. Chapter 3 deals with a more general problem of forecasting time series under structural change and long memory noise. Presence of long memory in the data is often easily confused with structural change. Wrongly account- ing for one when the other is present may lead to serious forecast failure. In our search for a forecast method that can perform reliably in presence of both features we extend the recent work of Giraitis et al. (2013). A forecast strategy with data-dependent discounting is adopted and typical robust-to- structural-change methods such as rolling window regression, forecast averag- ing and exponentially weighted moving average methods are exploited. We provide detailed theoretical analyses of forecast optimality by considering cer- tain types of structural changes and various degrees of long range dependence in noise. An extensive Monte Carlo study and empirical application to many UK time series ensure usefulness of adaptive forecast methods.

Despite significant regulatory and academic interest in sell-side analyst forecasts and an extensive literature demonstrating the impact of teamwork in general, we lack evidence of the effect of teamwork on analyst forecasts. In 2005 analyst teams issued nearly three-fourths of analyst reports for a sample of 89 large, heavily followed companies. Over a twelve-year period 86 of those companies had more reports issued by analyst teams than by individual analysts. Using a hand-collected sample of more than 17,000 analyst reports, I document that forecasts issued by analyst teams systematically differ from the forecasts of individual analysts in ways predicted by team literature. I find that prior to the year 2000 analyst teams issue forecasts that are less accurate and more biased than forecasts issued by individual analysts. Beginning in 2000, the relative benefit of analyst teamwork strengthens, consistent with changes due to Regulation Fair Disclosure, brokerage closures, and other regulatory interventions. In addition I find that, within company-year, team-issued forecasts are less pessimistically biased but not less optimistically biased than the forecasts issued by individual analysts. Lastly, the benefits of teamwork vary with the size of the team and over the life of the team, following an inverted u-shaped pattern. My results inform regulators as they consider factors that impact analyst forecast accuracy and bias.

Real-time wave forecasts are critical to a variety of coastal and offshore opera- tions. NOAA’s global wave forecasts, at present, do not extend into many coastal regions of interest. Even after more than two decades of the historical Exxon Valdez disaster, Cook Inlet (CI) and Prince William Sound (PWS) are regions that suffer from a lack of accurate wave forecast information. This dissertation develops high- resolution integrated wave forecasting schemes for these regions in order to meet the critical requirements associated with shipping, commercial and sport fishing vessel safety, and oil spill response. This dissertation also performs a detailed qualitative and quantitative assessment of the impact of various forcing functions on wave pre- dictions, and develops maps showing extreme variations in significant wave heights (SWHs). For instance, it is found that the SWH could vary by as much as 1 m in the northern CI region in the presence of currents (hence justifying the need for integration of the wave model with a circulation model). Such maps can be useful for several engineering operations, and could also serve as guidance tool as to what can be expected in certain regions. Aside from the system development, the issue of forecast reliability is also addressed for PWS region in the context of the associated uncertainty which confronts the manager of engineering operations or other planners. For this purpose, high-resolution 36-h daily forecasts of SWHs are compared with measurements from buoys and satellites for about a year. The results show that 70% of the peak SWHs in the range 5-8 m were predicted with an accuracy of 15% or less for a forecast lead time of 9 h. On average, results indicate 70% or greater likelihood of the prediction falling within a tolerance of ±(1*RMSE) for all lead times. This analysis could not be performed for CI due to lack of data sources.

This paper considers the high-moments and uses the skew generalized error distribution (SGED) to explain the financial market data which have leptokurtic, fat-tailed and skewness. And we compare performance with the commonly used symmetrical distribution model such as normal distribution, student¡¦s t distribution and generalized error distribution (GED). To research when returns of asset have leptokurtic and fat-tailed phenomena, what model has better predictive power for volatility forecasting? The empirical procedure is as follows: First step, make the descriptive statistics of raw data, and know that the GARCH effect should be considered, followed by selecting the optimal order of ARMA-GARCH. The second steps, make the parameter estimations of full-sample, and pick up the best model. Finally, forecast out-of-sample volatility for 1-day, 2-day, 5-day, 10-day and 20-day respectively, not only use different loss function to measure the performance, but also use DM test to compare the relative predictive power of the models under the different error distribution.