No description available.

A common problem in the area of non-linear function optimisation is that of not being able to guarantee finding the global optimum of the function in a feasible time especially when local optima exist. This problem applies to various areas of heuristic search. One of these areas concerns standard training techniques for feedforward neural networks. The element of heuristic search consists of attempting to find a neural weight state corresponding to the lowest training error. This problem may be termed the local minimum problem. The local minimum problem is addressed for feedforward neural networks. This is done by first establishing the conditions under which local minimum interference for the training process is to be expected. A target based approach to subgoal chaining in supervised learning is then investigated. This is a method to improve travel for neural networks by directing it more precisely through local subgoals than may be achieved through a more distant goal. It is shown however that linear subgoal chains are not sufficient to overcome the local minimum problem. Two novel training techniques are presented which use non-linear subgoal chains and are examined for their capability to address the local minimum problem. It is found that attempting to target a neural network to do something it cannot may lead to suboptimal training. It is also found that targeting a network to do something it is capable of generally leads to successful training. A novel system is presented which is designed to create optimal realisable targets for unrealisable goals. This allows neural networks to subsequently achieve the optimal weight state through a sufficiently powerful training method such as subgoal chaining. The results are shown to be consistent with the theoretical expectations.

The creation of anatomically correct three-dimensional joints for the simulation of humans is a complex process, a key difficulty being the correction of invalid joint configurations to the nearest valid alternative. Personalised models based on individual joint mobility are in demand in both animation and medicine [1]. Medical models need to be highly accurate animated models less so, however if either are to be used in a real time environment they must have a low temporal cost (high performance). This work briefly explores Support Vector Machine neural networks as joint configuration classifiers that group joint configurations into invalid and valid. A far more detailed investigation is carried out into the use of topologically evolved feed forward neural networks for the generation of appropriately proportioned corrective components which when applied to an invalid joint configuration result in a valid configuration and the same configuration if the original configuration was valid. Discontinuous vector fields were used to represent constraints of varying size, dimensionality and complexity. This culminated in the creation corrective quaternion constraints represented by discontinuous vector fields, learned by topologically evolved neural networks and trained via the resilient back propagation algorithm. Quaternion constraints are difficult to implement and although alternative methods exist [2-6] the method presented here is superior in many respects. This method of joint constraint forms the basis of the contribution to knowledge along with the discovery of relationships between the continuity and distribution of samples in quaternion space and neural network performance. The results of the experiments for constraints on the rotation of limb with regular boundaries show that 3.7 x lO'Vo of patterns resulted in errors greater than 2% of the maximum possible error while for irregular boundaries 0.032% of patterns resulted in errors greater than 7.5%.

The thesis is written in chapter form. Chapter 1 describes some of the history of neural networks and its place in the field of artificial intelligence. It indicates the biological basis from which neural network approximation are made. Chapter 2 describes the properties of neural networks and their uses. It introduces the concepts of training and learning. Chapters 3, 4, 5 and 6 show the perceptron and adaline in feedforward and recurrent networks particular reference is made to regression substitution by "group method data handling. Networks are chosen that explain the application of neural networks in classification, association, optimization and self organization. Chapter 7 addresses the subject of practical inputs to neural networks. Chapter 8 reviews some interesting recent developments. Chapter 9 reviews some ideas on the future technology for neural networks. Chapter 10 gives a listing of some neural network types and their uses. Appendix A gives some of the ideas used in portfolio selection for the Johannesburg Stock Exchange. / Computing / M. Sc. (Operations Research)

Neural networks need to be able to guarantee their intrinsic generalisation abilities if they are to be used reliably. Mitchell's concept and version spaces technique is able to guarantee generalisation in the symbolic concept-learning environment in which it is implemented. Generalisation, according to Mitchell, is guaranteed when there is no alternative concept that is consistent with all the examples presented so far, except the current concept, given the bias of the user. A form of bidirectional convergence is used by Mitchell to recognise when the no-alternative situation has been reached. Mitchell's technique has problems of search and storage feasibility in its symbolic environment. This thesis aims to show that by evolving the technique further in a neural environment, these problems can be overcome. Firstly, the biasing factors which affect the kind of concept that can be learned are explored in a neural network context. Secondly, approaches for abstracting the underlying features of the symbolic technique that enable recognition of the no-alternative situation are discussed. The discussion generates neural techniques for guaranteeing generalisation and culminates in a neural technique which is able to recognise when the best fit neural weight state has been found for a given set of data and topology.

The application of neural networks in the area of signal processing is examined. Two major areas are identified and suitable neural networks are developed. In the first area, neural networks are used as a tool for the design of digital filters. In the second area, neural networks are used for processing bathymetric data. The field of artificial neural networks is first introduced with an emphasis on Hopfield networks. The optimizing capabilities of such networks are noted. Based on these networks, a feedback neural network is developed for the design of 1-D finite-duration impulse response (FIR) filters on the basis of given amplitude responses. A suitable cost function is formulated first and an associated network is developed. This work is then extended to the design of two more networks for the design of FIR filters based on given amplitude and phase responses and prescribed specifications. The idea is extended to the design of 2-D FIR filters. Two networks are presented for designing 2-D FIR filters on the basis of a given amplitude response and prescribed specifications. The design of 1-D infinite-duration impulse response (IIR) filters is studied next and two networks are developed. The first one is to design filters with prescribed specifications in the magnitude-squared domain. The other network designs IIR filters for a given frequency response. A network for designing equiripple 1-D FIR filters based on the weighted least-squares technique is presented next. A new updating algorithm is developed for this network. Two different neural networks are proposed for classifying lidar waveforms into various categories. A single-layer network is developed for classifying lidar waveforms representing milt of varied densities. A fast version of the supervised learning algorithm is presented. A threshold term is also introduced in the recall phase to give the user flexibility to accept or reject any waveform. A two-stage, multi-layer network is presented next which uses waveform characteristics to assign a signature number to the waveform. This network extracts various ocean parameters from the waveforms as well. The issue of implementing the feedback neural network is addressed next. Basic building blocks for implementing such networks are identified and a network is constructed from circuits existing in the literature. The network is simulated in Cadence using 0.8 μ BICMOS technology. The results show that these networks have a high potential to be implemented in analog VLSI for real-time signal processing. / Graduate

M.Ing. (Electrical And Electronic Engineering) / Please refer to full text to view abstract

M.Ing. / A great deal of work has been done in the field of hexapodous autonomous agents. However, in this dissertation the locomotion of a more complex organism - the octopod - will be studied. Biological neural behaviour will present a basis for the leg controllers, while classic backpropagation networks will be used to implement pattern generators. Full simulation of the biological scorpion leg will be implemented, thus a simulated leg consisting of six joint angles with 6 degrees of freedom. Simple locomotion on a flat substrate will be considered. In this dissertation the scorpion will be used as basis of simulation mainly due to the interesting leg architecture and intricate locomotory patterns during locomotion, hunting and burrowing. The locomotory models developed here may be modified to facilitate other terrestial octopodous agents.

M. Ing. / This dissertation describes the development of a feedforwa.rd neural network, trained by means of an accelerated backpropagation algorithm, used for the short term load forecasting on real world data. It is argued that the new learning algorithm. I-Prop, - is a faster training - algorithm due to the fact that the learning rate is optimally predicted and changed according to a more efficient formula (without the need for extensive memory) which speeds up the training process. The neural network developed was tested for the month of December 1994, specifically to test the artificial neural network's ability to correctly predict the load during a Public Holiday, as well as the change over from Public Holiday to 'Normal' working day. In conclusion, suggestions are made towards further research in the improvement of the I-Prop algorithm as well as improving the load forecasting technique implemented in this dissertation.

Spread F is a phenomenon of the ionosphere in which the pulses returned from the ionosphere are of a much greater duration than the transmitted ones. The occurrence of spread F can be predicted using the technique of Neural Networks (NNs). This thesis presents the development and evaluation of NN based models (two single station models and a regional model) for predicting the occurrence of spread F over selected stations within the Brazilian sector. The input space for the NNs included the day number (seasonal variation), hour (diurnal variation), sunspot number (measure of the solar activity), magnetic index (measure of the magnetic activity) and magnetic position (latitude, magnetic declination and inclination). Twelve years of spread F data measured during 1978 to 1989 inclusively at the equatorial site Fortaleza and low latitude site Cachoeira Paulista are used in the development of an input space and NN architecture for the NN models. Spread F data that is believed to be related to plasma bubble developments (range spread F) were used in the development of the models while those associated with narrow spectrum irregularities that occur near the F layer (frequency spread F) were excluded. The results of the models show the dependency of the probability of spread F as a function of local time, season and latitude. The models also illustrate some characteristics of spread F such as the onset and peak occurrence of spread F as a function of distance from the equator. Results from these models are presented in this thesis and compared to measured data and to modelled data obtained with an empirical model developed for the same purpose.