Thesis (MBA)--Stellenbosch University, 2006. / Although most businesses require accurate sales forecasts in order to survive and to be successful, very little attention has been devoted to examine how sales forecasting processes should be managed, and the behavioural factors associated with the management of forecasting. Sales forecasting activities and research have by and large concentrated on the techniques or on the systems used, rather than on the forecasting management philosophy, which considers the organisational, procedural, and personnel aspects of the process. Both forecasting modelling and IT systems form the basis for the forecasting process, but the third element, namely the organisation, is potentially the most important one. Researchers have argued that improvements in this area could have a greater impact on the level of forecasting accuracy than improvements with regard to other aspects. After developing predetermined forecasting standards and principles, an audit on the author's organisation was conducted. This revealed that no formal forecasting --- existed, and that a number of business practices were in effect contaminating procedures and possibly affecting the integrity of the data. Very little forecasting knowledge existed, sales were predicted very sporadically, and simple averaging techniques were adopted. Life cycles of products, trends, seasonality or any other cyclical activity were never modelled. This obviously resulted in a very poor level of forecast accuracy, affecting a number of business activities. A decision was made to research the topic of forecasting management, develop a best practice model, and apply it to the organisation. The best practice model was based predominantly on the research work of Armstrong and Mentzer. This model requires the forecasting process to be developed in two specific phases, namely a strategic phase, in which the forecast is aligned to the organisation, the internal processes and the people, and the operational phase, in which more tangible aspects of the forecasting process are identified and constructed. This new forecasting approach and a dedicated forecasting software programme were successfully implemented, improving the overall accuracy level of the forecast.

Thesis (MBA)--Stellenbosch University, 1985. / Aangesien historiese data geredelik beskikbaar was, is 'n kwantitatiewe vooruitskattingsmetode gebruik met die doel om gebeure in die verlede te bestuur. Sodoende kon die onderliggende struktuur van die data beter begryp word en daarom kon 'n model daargestel word om die nodige inligting te verskaf vir bestuursbesluitneming. Die klassieke vermenigvuldigende tydreeks is gebruik om die toekomstige verkope van Stodels Nurseries (Edms.) Bpk. te projekteer. Aangesien die maatskappy se verkope onderhewig is aan hewige seisoenskommelings, is kontantvloeibeplanning van kardinale belang vir die finansiele bestuur van die maatskappy.

The increasing frequency of sporadic weather patterns in the last decade, especially major winter storms, demands improvements in current weather forecasting techniques. Recently, there are growing interests in stratospheric forecasting because of its potential enhancements of weather forecasts. The dominating factors of northern hemisphere wintertime variation of the general circulation in the stratosphere is a phenomenon called stratospheric sudden warming (SSW) events. It is shown in multiple studies that SSW and cosmic ray muon flux variations are strongly correlated with the effective atmospheric temperature changes, which suggests that cosmic ray detectors could be potentially used as meteorological applications, especially for monitoring SSW events. A method for determining the effective temperature with cosmic ray flux measurements is studied in this work by using statistical modeling techniques, such as k-fold cross validation and partial least square regression. This method requires the measurement of the vertical profile of the atmospheric temperature, typically measured by radiosonde, for training the model. In this study, cosmic ray flux measured in Atlanta and Yakutsk are chosen for demonstrating this novel technique. The results of this study show the possibility of realtime monitoring on effective temperature by simultaneous measurement of cosmic ray muon and neutron flux. This technique can also be used for studying the historical SSW events using the past world wide cosmic ray data.

published_or_final_version / Business Administration / Master / Master of Business Administration

published_or_final_version / Economics and Finance / Doctoral / Doctor of Philosophy

published_or_final_version / Education / Master / Master of Science in Information Technology in Education

published_or_final_version / abstract / Statistics and Actuarial Science / Master / Master of Philosophy

published_or_final_version / Economics and Finance / Doctoral / Doctor of Philosophy

Seasonal influenza epidemics are a major public health concern, causing three to five million cases of severe illness and about 250,000 to 500,000 deaths worldwide. Given the unpredictability of these epidemics, hospitals and health authorities are often left unprepared to handle the sudden surge in demand. Hence early detection of disease activity is fundamental to reduce the burden on the healthcare system, to provide the most effective care for infected patients and to optimize the timing of control efforts. Early detection requires reliable forecasting methods that make efficient use of surveillance data. We developed a dynamic Bayesian estimator to predict weekly hospitalizations due to influenza related illnesses in the state of Texas. The prediction of peak hospitalizations using our model is accurate both in terms of number of hospitalizations and the time at which the peak occurs. For 1-to 8 week predictions, the predicted number of hospitalizations was within 8% of actual value and the predicted time of occurrence was within a week of actual peak. / text

The purpose of this study is to investigate the daily return behavior of underlying common stocks in the period surrounding the option expiration date. A second purpose is to determine the variables that may be causing the differential capital market effect across firms. The hypothesis of a negative return effect in the expiration week followed by a positive effect in the subsequent week is tested first. It is shown that this pattern should be expected due to the enhanced opportunity for and profitability of position unwinding, arbitrage and manipulation activity as the expiration date approached. The study period covers 32 expiration periods from 1978 through 1981 and involves a sample of 138 underlying stocks. The study employs the market model for generating abnormal returns on a daily basis. The results support the hypothesis and in particular show that the most significant negative return behavior occurs on Thursday and Friday of the expiration week. The second phase of the study correlates, via a cross-sectional multiple regression model, the suggested expiration induced events of position unwinding, arbitrage and manipulation activities with the return behavior of the underlying stocks. It is hypothesized that those common stocks which exhibit the greatest negative returns in the expiration week are those stocks and related call options that are most heavily involved in position unwinding, arbitrage and manipulation activities. Trading volume in both the underlying stock and the options is suggested as a surrogate for these three activities. Therefore, volume is negatively related to underlying stock returns. Two additional explanatory variables of the expiration week returns are included in the regression model. A negative relationship is hypothesized if options are dually listed and a positive relationship if puts are traded. The results of the tests generally support these hypothesized functional relationships. The study concludes that, although significant abnormal returns and explanatory variables are found, the magnitudes are probably not large enough to profitably exploit after paying transaction and search costs. As puts trading appears to offset the market inefficiencies caused by call option trading, the concern of regulators that options trading unduly affects stock prices seems unwarranted.