Accelerated degradation testing is widely accepted in competitive industries. As there is no longer the need to test till failures, there are tremendous cost and time benefits on fully capitalizing on such a testing regime. Consequently, this research has aimed for better understanding of the relationship between design and degradation using the degradation data. Existing work in the literature uses the degradation data to improve the reliability of products. The majority of techniques, however, are centred on statistical experimental methods. For problems with increasing complexities such as large multivariable data set, non-linear interactions and dynamic varying processes, conventional methods cannot resolve the problem efficiently. Furthermore, it can not provide the adequate modelling mechanism to learn from the degradation data autonomously for describing the relationship between the design parameters and degradation. Artificial neural network is widely used for complex problems in the literature. This thesis proposes and demonstrates the neural network modelling methodology into capturing the non parametric relationship between design and degradation. / The development of a neural network consists of data preparation, network design and training and testing. This thesis presents a comprehensive description on the data generation and acquisition process. More specifically, the physical tests, experimental designs, equipment configurations, data acquisitions systems and algorithms are elaborated. Single hidden layer multilayered perceptrons are found to be the most suitable network architectures for the problem domains. Detailed descriptions of the training and testing process in determining the suitable number of hidden neurons sufficient for the problem are provided. / In summary, the neural network modelling methodology is demonstrated for the particular problem domain. As a result of the work in this thesis, two models of different practical significance are developed and compiled as Windows executables for predicting material performances. / Thesis (MEng(ManufacturingEngineering)--University of South Australia, 2006.

The area of Agent Technology is a relatively new area and one that has generated great interest within computing circles. The concept of software identities that have the intelligence to perform some of the tasks that humans perform has great potential. Such software identities are slowly making their way into many complex systems ranging from air traffic control systems to web search engines, and removing the direct need for the 'human in the loop'. This type of software that has the ability or intelligence to perform some of the tasks that would otherwise be performed by humans has been coined the term 'Intelligent Agents'. / In many of these agent based systems there may be a series of agents that are operating together to achieve common goals. Such systems are often referred to as a multiagent system. Agents within a multiagent system, usually share information and knowledge between each other in a collaborative manner to achieve a common goal. The information that is shared around is normally learned knowledge. / For the study conducted for this Master's project, collaborative agent learning is examined. The main emphasis of this study is to research how agents can learn. This learned knowledge can then be shared amongst other agents in a system to achieve a common goal. To investigate this, a case study was developed, where the aim was to develop an Automatic Target Recognition (ATR) system containing two main types of agents- a Detection Agent and a Recognition Agent. These agents would have specific roles within the system; however both would be trained to identify the same target images. These agents would then collaborate with each other to positively identify these target images when presented with masses of data in the form of test images as input into the ATR system. / The intelligent and learning aspects of these agents are implemented using Artificial Neural Networks. Two different Artificial Neural Network techniques are used for the two agents. This thesis contains a detailed discussion on Agent Technology, Artificial Neural Networks, the ATR system and descriptions of how the Detection and Recognition Agents were developed and tested. Experimental results are also presented with a discussion on the overall success of the ATR system, the results obtained, problems encountered and future directives for this research. / Thesis (MEng(ElectronicsEngineering))--University of South Australia, 2005.

Traditional Automatic Target Recognition (ATR) Systems often fail when faced with complex recognition tasks involving noise, clutter, and complexity. This work is concerned with implementing a real time, vision based ATR system using an Artificial Neural Network (ANN) to overcome some of the shortcomings of traditional ATR systems. The key issues of this work are vision, pattern recognition and artificial neural networks. The ANN presented in this thesis is inspired by Prof. Stephen Grossberg's work in Adaptive Resonance Theory (ART) and neurophysiological data on the primate brain. An ANN known as Selective Attention Adaptive Resonance Theory (SAART) (Lozo, 1995, 1997) forms the basis of this work. SAART, which is based on Grossberg's ART, models the higher levels of visual processing in the primate brain to provide an ATR system capable of learning and recognising targets in cluttered and complex backgrounds. This thesis contributes an extension to the SAART model to allow a degree of tolerance to imperfections including distortion, changes in size, orientation, or position. In addition to this extension, it is also demonstrated how modulated neural layers can be used for image filtering. A possible extension of the architecture for multi-sensory environments is proposed as a foundation for future research. / Thesis (MEng)--University of South Australia, 2000

Bibliography: leaves 328-352. / xxxi, 352 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / This thesis proposes a neural theory, Selective Attention Adaptive Resonance Theory, and a neuro-engineered solution to selective visual attention, memory guided processing and illumination invariant recognition of complete (unoccluded), but distorted 2D shapes of 3D objects in cluttered visual images. / Thesis (Ph.D.)--University of Adelaide, Dept. of Electrical and Electronic Engineering, 1997?

Bibliography: leaves 125-131. / xi, 131 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / This thesis examines the performace of optimisation networks. The main objectives are to determine if there exist any factors which limit the solution quality that may be achieved with optimisation networks, to determine the reasons for any such limitations and to suggest remedies for them. / Thesis (Ph.D.)--University of Adelaide, Dept. of Electrical and Electronic Engineering, 1996

Bibliographical references: p. 225-234. / xxiv, 285 p. : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Thesis (Ph.D.)--University of Adelaide, Dept. of Electrical and Electronic Engineering, 1998

Includes bibliographical references (leaves 225-241). / xxviii, 241 leaves : ill. ; 30 cm. / Title page, contents and abstract only. The complete thesis in print form is available from the University Library. / Proposes an architectural framework based on neural networks for visual scene analysis with attentional mechanisms. / Thesis (Ph.D.)--Adelaide University, Dept. of Electrical and Electronic Engineering, 2001

Traditional Automatic Target Recognition (ATR) Systems often fail when faced with complex recognition tasks involving noise, clutter, and complexity. This work is concerned with implementing a real time, vision based ATR system using an Artificial Neural Network (ANN) to overcome some of the shortcomings of traditional ATR systems. The key issues of this work are vision, pattern recognition and artificial neural networks. The ANN presented in this thesis is inspired by Prof. Stephen Grossberg's work in Adaptive Resonance Theory (ART) and neurophysiological data on the primate brain. An ANN known as Selective Attention Adaptive Resonance Theory (SAART) (Lozo, 1995, 1997) forms the basis of this work. SAART, which is based on Grossberg's ART, models the higher levels of visual processing in the primate brain to provide an ATR system capable of learning and recognising targets in cluttered and complex backgrounds. This thesis contributes an extension to the SAART model to allow a degree of tolerance to imperfections including distortion, changes in size, orientation, or position. In addition to this extension, it is also demonstrated how modulated neural layers can be used for image filtering. A possible extension of the architecture for multi-sensory environments is proposed as a foundation for future research. / Thesis (MEng)--University of South Australia, 2000

[Formulae and special characters can only be approximated here. Please see the pdf version of the Abstract for an accurate reproduction.] Artificial neural networks are universal function approximators that represent functions subsymbolically by weights, thresholds and network topology. Naturally, the representation remains the same regardless of the problem domain. Suppose a network is applied to a symbolic domain. It is difficult for a human to dynamically construct the symbolic function from the neural representation. It is also difficult to retrain networks on perturbed training vectors, to resume training with different training sets, to form a new neuron by combining trained neurons, and to reason with trained neurons. Even the original training set does not provide a symbolic representation of the function implemented by the trained network because the set may be incomplete or inconsistent, and the training phase may terminate with residual errors. The symbolic information in the network would be more useful if it is available in the language of the problem domain. Algorithms that translate the subsymbolic neural representation to a symbolic representation are called extraction algorithms. I argue that extraction algorithms that operate on single-output, layered feedforward networks are sufficient to analyse the class of multiple-output networks with arbitrary connections, including recurrent networks. The translucency dimensions of the ADT taxonomy for feedforward networks classifies extraction approaches as pedagogical, eclectic, or decompositional. Pedagogical and eclectic approaches typically use a symbolic learning algorithm that takes the network’s input-output behaviour as its raw data. Both approaches construct a set of input patterns and observe the network’s output for each pattern. Eclectic and pedagogical approaches construct the input patterns respectively with and without reference to the network’s internal information. These approaches are suitable for approximating the network’s function using a probably-approximately-correct (PAC) or similar framework, but they are unsuitable for constructing the network’s complete function. Decompositional approaches use internal information from a network more directly to produce the network’s function in symbolic form. Decompositional algorithms have two components. The first component is a core extraction algorithm that operates on a single neuron that is assumed to implement a symbolic function. The second component provides the superstructure for the first. It consists of a decomposition rule for producing such neurons and a recomposition rule for symbolically aggregating the extracted functions into the symbolic function of the network. This thesis makes contributions to both components for Boolean extraction. I introduce a relatively efficient core algorithm called WSX based on a novel Boolean form called BvF. The algorithm has a worst case complexity of O(2 to power of n divided by the square root of n) for a neuron with n inputs, but in all cases, its complexity can also be expressed as O(l) with an O(n) precalculation phase, where l is the length of the extracted expression in terms of the number of symbols it contains. I extend WSX for approximate extraction (AWSX) by introducing an interval about the neuron’s threshold. Assuming that the input patterns far from the threshold are more symbolically significant to the neuron than those near the threshold, ASWX ignores the neuron’s mappings for the symbolically input patterns, remapping them as convenient for efficiency. In experiments, this dramatically decreased extraction time while retaining most of the neuron’s mappings for the training set. Synthetic decomposition is this thesis’ contribution to the second component of decompositional extraction. Classical decomposition decomposes the network into its constituent neurons. By extracting symbolic functions from these neurons, classical decomposition assumes that the neurons implement symbolic functions, or that approximating the subsymbolic computation in the neurons with symbolic computation does not significantly affect the network’s symbolic function. I show experimentally that this assumption does not always hold. Instead of decomposing a network into its constituent neurons, synthetic decomposition uses constraints in the network that have the same functional form as neurons that implement Boolean functions; these neurons are called synthetic neurons. I present a starting point for constructing synthetic decompositional algorithms, and proceed to construct two such algorithms, each with a different strategy for decomposition and recomposition. One of the algorithms, ACX, works for networks with arbitrary monotonic transfer functions, so long as an inverse exists for the functions. It also has an elegant geometric interpretation that leads to meaningful approximations. I also show that ACX can be extended to layered networks with any number of layers.

Thesis (Ph. D.)--West Virginia University, 2007. / Title from document title page. Document formatted into pages; contains xiii, 200 p. : ill. (some col.). Includes abstract. Includes bibliographical references (p. 143-151).