A research report submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master of Science, Johannesburg, 30 March 2018. / Accuracy of the load forecasts is very critical in the power system industry, which is the lifeblood of the global economy to such an extent that its art-of-the-state management is the focus of the Short-Term Load Forecasting (STLF) models. In the past few years, South Africa faced an unprecedented energy management crisis that could be addressed in advance, inter alia, by carefully forecasting the expected load demand. Moreover, inaccurate or erroneous forecasts may result in either costly over-scheduling or adventurous under-scheduling of energy that may induce heavy economic forfeits to power companies. Therefore, accurate and reliable models are critically needed. Traditional statistical methods have been used in STLF but they have limited capacity to address nonlinearity and non-stationarity of electric loads. Also, such traditional methods cannot adapt to abrupt weather changes, thus they failed to produce reliable load forecasts in many situations. In this research report, we built a STLF model using Artificial Neural Networks (ANNs) to address the accuracy problem in this field so as to assist energy management decisions makers to run efficiently and economically their daily operations. ANNs are a mathematical tool that imitate the biological neural network and produces very accurate outputs. The built model is based on the Multilayer Perceptron (MLP), which is a class of feedforward ANNs using the backpropagation (BP) algorithm as its training algorithm, to produce accurate hourly load forecasts. We compared the MLP built model to a benchmark Seasonal Autoregressive Integrated Moving Average with Exogenous variables (SARIMAX) model using data from Eskom, a South African public utility. Results showed that the MLP model, with percentage error of 0.50%, in terms of the MAPE, outperformed the SARIMAX with 1.90% error performance. / LG2018

No description available.

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Electrical and Electronic Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Mechanical Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Mechanical Engineering / Doctoral / Doctor of Philosophy

published_or_final_version / Industrial and Manufacturing Systems Engineering / Doctoral / Doctor of Philosophy

Six alkalinity and oxidation classes of fresh igneous rocks were correlated with trace elements in rock chip samples from temporally and spatially associated ore deposits. Learning vector quantization and back-propagation artificial neural networks correctly classified 100 percent of whole rock oxides and 99 percent of mineralized samples; discriminant analysis correctly classified 96 and 83 percent, respectively. The high degree of correlation between chemistries of igneous rocks and related mineralization implies genetic links between magmatic processes or sources and the ore deposits studied. The petrochemical classification was evaluated by assigning 43 deposits to classes defined on eight variation diagrams, training neural networks to classify analyses of 569 igneous and 887 mineralized samples, and testing the networks on their ability to classify new data. Whole rock analyses were obtained from mining districts in which trace element geochemistry was also available. Half the data was eliminated using five alteration filter graphs. The K₂O and Fe₂O₃/FeO versus SiO₂ diagrams and iron mineralogy best defined alkalinity and oxidation classes. Neural networks trained with 90, 80, 70, or 50 percent of the samples correctly classified 81 to 100 percent of randomly withheld data. SiO₂/K₂O ratios of alkali-calcic igneous rocks are 14-20 and of calc-alkalic 20-30. Fe₂O₃/FeO ratios are >0.8 with abundant magnetite and sphene for oxidized, 0.5-1.2 with magnetite, sphene, and rare ilmenite for weakly oxidized, and <0.6 with ilmenite only in reduced subclasses. Lead-zinc-silver deposits as at Tombstone and Tintic are related to oxidized alkali-calcic igneous rocks. Polymetallic lead-zinc-copper-tin-silver deposits, such as Santa Eulalia and Tempiute, Nevada, are associated with weakly oxidized alkali-calcic rocks. Tin-silver deposits of Llallagua and Potosi are correlated with reduced alkali-calcic intrusives. Porphyry copper deposits as at Ray and Sierrita are connected with oxidized calc-alkalic plutons. Gold-rich porphyry copper deposits, such as Copper Canyon and Morenci are linked to weakly oxidized calc-alkalic plutons. Disseminated gold deposits, such as Chimney Creek, Nevada, are temporally and chemically correlated with reduced calc-alkalic igneous rocks, although physical connections between plutons and Carlin-type deposits remain unconfirmed. Magma series classification and neural networks have profound applications and implications to exploration, alteration and zoning studies, and metallogenesis.

An optical computer which performs the classification of an input object pattern into one of two learned classes is designed and demonstrated. The classifier is an optical implementation of a neural network model of computation featuring learning, self-organization, and decision-making competition. Neural computation is discussed including models for learning networks and motivation for optical implementation. A discussion of photorefractive crystal holographic storage and adaptation is presented followed by experimental results of writing and erasing gratings in several different crystals. The optical network features a photorefractive crystal to store holographic interconnection weights and an opto-electronic circuit to provide a means of competitive decision making and feedback. Results of the optical learning network and its operation as an associative memory are followed by extensions of the architecture to allow improved performance and greater flexibility.

Huge amounts of data are being collected as a result of the increased use of mobile telecommunications. Insight into information and knowledge derived from these databases can give operators a competitive edge in terms of customer care and retention,<br /> marketing and fraud detection. One of the strategies for fraud detection checks for signs of questionable changes in user behavior. Although the intentions of the mobile phone users cannot be observed, their intentions are reflected in the call data which define usage patterns. Over a period of time, an individual phone generates a large pattern of use. While call data are recorded for subscribers for billing purposes, we are making no prior assumptions about the data indicative of fraudulent call patterns, i.e. the calls made for billing purpose are unlabeled. Further analysis is thus, required to be able to isolate fraudulent usage. An unsupervised learning algorithm can analyse and cluster call patterns for each subscriber in order to facilitate the fraud detection process.<br /> <br /> This research investigates the unsupervised learning potentials of two neural networks for the profiling of calls made by users over a period of time in a mobile telecommunication network. Our study provides a comparative analysis and application of Self-Organizing Maps (SOM) and Long Short-Term Memory (LSTM) recurrent neural networks algorithms to user call data records in order to conduct a descriptive data mining on users call patterns.<br /> <br /> Our investigation shows the learning ability of both techniques to discriminate user call patterns / the LSTM recurrent neural network algorithm providing a better discrimination than the SOM algorithm in terms of long time series modelling. LSTM discriminates different types of temporal sequences and groups them according to a variety of features. The ordered features can later be interpreted and labeled according to specific requirements of the mobile service provider. Thus, suspicious call behaviours are isolated within the mobile telecommunication network and can be used to to identify fraudulent call patterns. We give results using masked call data<br /> from a real mobile telecommunication network.