The 'moving targets' training algorithm

A simple method for training the dynamical behavior of a neural network is derived. It is applicable to any training problem in discrete-time networks with arbitrary feedback. The algorithm resembles back-propagation in that an error function is minimized using a gradient-based method, but the optimization is carried out in the hidden part of state space either instead of, or in addition to weight space. Computational results are presented for some simple dynamical training problems, one of which requires response to a signal 100 time steps in the past.

Using ancillary statistics in on-line learning algorithms

Neural networks are usually curved statistical models. They do not have finite dimensional sufficient statistics, so on-line learning on the model itself inevitably loses information. In this paper we propose a new scheme for training curved models, inspired by the ideas of ancillary statistics and adaptive critics. At each point estimate an auxiliary flat model (exponential family) is built to locally accommodate both the usual statistic (tangent to the model) and an ancillary statistic (normal to the model). The auxiliary model plays a role in determining credit assignment analogous to that played by an adaptive critic in solving temporal problems. The method is illustrated with the Cauchy model and the algorithm is proved to be asymptotically efficient.

Incorporating curvature information into on-line learning

We analyse the dynamics of a number of second order on-line learning algorithms training multi-layer neural networks, using the methods of statistical mechanics. We first consider on-line Newton's method, which is known to provide optimal asymptotic performance. We determine the asymptotic generalization error decay for a soft committee machine, which is shown to compare favourably with the result for standard gradient descent. Matrix momentum provides a practical approximation to this method by allowing an efficient inversion of the Hessian. We consider an idealized matrix momentum algorithm which requires access to the Hessian and find close correspondence with the dynamics of on-line Newton's method. In practice, the Hessian will not be known on-line and we therefore consider matrix momentum using a single example approximation to the Hessian. In this case good asymptotic performance may still be achieved, but the algorithm is now sensitive to parameter choice because of noise in the Hessian estimate. On-line Newton's method is not appropriate during the transient learning phase, since a suboptimal unstable fixed point of the gradient descent dynamics becomes stable for this algorithm. A principled alternative is to use Amari's natural gradient learning algorithm and we show how this method provides a significant reduction in learning time when compared to gradient descent, while retaining the asymptotic performance of on-line Newton's method.

Equilibration through local information exchange in networks

We study the equilibrium states of energy functions involving a large set of real variables, defined on the links of sparsely connected networks, and interacting at the network nodes, using the cavity and replica methods. When applied to the representative problem of network resource allocation, an efficient distributed algorithm is devised, with simulations showing full agreement with theory. Scaling properties with the network connectivity and the resource availability are found. © 2006 The American Physical Society.

Outlier detection with partial information:Application to emergency mapping

This paper, addresses the problem of novelty detection in the case that the observed data is a mixture of a known 'background' process contaminated with an unknown other process, which generates the outliers, or novel observations. The framework we describe here is quite general, employing univariate classification with incomplete information, based on knowledge of the distribution (the 'probability density function', 'pdf') of the data generated by the 'background' process. The relative proportion of this 'background' component (the 'prior' 'background' 'probability), the 'pdf' and the 'prior' probabilities of all other components are all assumed unknown. The main contribution is a new classification scheme that identifies the maximum proportion of observed data following the known 'background' distribution. The method exploits the Kolmogorov-Smirnov test to estimate the proportions, and afterwards data are Bayes optimally separated. Results, demonstrated with synthetic data, show that this approach can produce more reliable results than a standard novelty detection scheme. The classification algorithm is then applied to the problem of identifying outliers in the SIC2004 data set, in order to detect the radioactive release simulated in the 'oker' data set. We propose this method as a reliable means of novelty detection in the emergency situation which can also be used to identify outliers prior to the application of a more general automatic mapping algorithm. © Springer-Verlag 2007.

Implementing an adaptive TCP fairness while exploiting 802.11e over wireless mesh networks

Purpose: The purpose of this paper is to investigate the use of 802.11e MAC to resolve the transmission control protocol (TCP) unfairness. Design/methodology/approach: The paper shows how a TCP sender may adapt its transmission rate using the number of hops and the standard deviation of recently measured round-trip times to address the TCP unfairness. Findings: Simulation results show that the proposed techniques provide even throughput by providing TCP fairness as the number of hops increases over a wireless mesh network (WMN). Research limitations/implications: Future work will examine the performance of TCP over routing protocols, which use different routing metrics. Other future work is scalability over WMNs. Since scalability is a problem with communication in multi-hop, carrier sense multiple access (CSMA) will be compared with time division multiple access (TDMA) and a hybrid of TDMA and code division multiple access (CDMA) will be designed that works with TCP and other traffic. Finally, to further improve network performance and also increase network capacity of TCP for WMNs, the usage of multiple channels instead of only a single fixed channel will be exploited. Practical implications: By allowing the tuning of the 802.11e MAC parameters that have previously been constant in 802.11 MAC, the paper proposes the usage of 802.11e MAC on a per class basis by collecting the TCP ACK into a single class and a novel congestion control method for TCP over a WMN. The key feature of the proposed TCP algorithm is the detection of congestion by measuring the fluctuation of RTT of the TCP ACK samples via the standard deviation, plus the combined the 802.11e AIFS and CWmin allowing the TCP ACK to be prioritised which allows the TCP ACKs will match the volume of the TCP data packets. While 802.11e MAC provides flexibility and flow/congestion control mechanism, the challenge is to take advantage of these features in 802.11e MAC. Originality/value: With 802.11 MAC not having flexibility and flow/congestion control mechanisms implemented with TCP, these contribute to TCP unfairness with competing flows. © Emerald Group Publishing Limited.

Visual data mining: integrating machine learning with information visualization

Today, the data available to tackle many scientific challenges is vast in quantity and diverse in nature. The exploration of heterogeneous information spaces requires suitable mining algorithms as well as effective visual interfaces. Most existing systems concentrate either on mining algorithms or on visualization techniques. Though visual methods developed in information visualization have been helpful, for improved understanding of a complex large high-dimensional dataset, there is a need for an effective projection of such a dataset onto a lower-dimension (2D or 3D) manifold. This paper introduces a flexible visual data mining framework which combines advanced projection algorithms developed in the machine learning domain and visual techniques developed in the information visualization domain. The framework follows Shneiderman’s mantra to provide an effective user interface. The advantage of such an interface is that the user is directly involved in the data mining process. We integrate principled projection methods, such as Generative Topographic Mapping (GTM) and Hierarchical GTM (HGTM), with powerful visual techniques, such as magnification factors, directional curvatures, parallel coordinates, billboarding, and user interaction facilities, to provide an integrated visual data mining framework. Results on a real life high-dimensional dataset from the chemoinformatics domain are also reported and discussed. Projection results of GTM are analytically compared with the projection results from other traditional projection methods, and it is also shown that the HGTM algorithm provides additional value for large datasets. The computational complexity of these algorithms is discussed to demonstrate their suitability for the visual data mining framework.

A constructive learning algorithm based on back-propagation

There are been a resurgence of interest in the neural networks field in recent years, provoked in part by the discovery of the properties of multi-layer networks. This interest has in turn raised questions about the possibility of making neural network behaviour more adaptive by automating some of the processes involved. Prior to these particular questions, the process of determining the parameters and network architecture required to solve a given problem had been a time consuming activity. A number of researchers have attempted to address these issues by automating these processes, concentrating in particular on the dynamic selection of an appropriate network architecture.The work presented here specifically explores the area of automatic architecture selection; it focuses upon the design and implementation of a dynamic algorithm based on the Back-Propagation learning algorithm. The algorithm constructs a single hidden layer as the learning process proceeds using individual pattern error as the basis of unit insertion. This algorithm is applied to several problems of differing type and complexity and is found to produce near minimal architectures that are shown to have a high level of generalisation ability.

Neural network enhanced self tuning adaptive control application for non-linear control of dynamic systems

The main theme of research of this project concerns the study of neutral networks to control uncertain and non-linear control systems. This involves the control of continuous time, discrete time, hybrid and stochastic systems with input, state or output constraints by ensuring good performances. A great part of this project is devoted to the opening of frontiers between several mathematical and engineering approaches in order to tackle complex but very common non-linear control problems. The objectives are: 1. Design and develop procedures for neutral network enhanced self-tuning adaptive non-linear control systems; 2. To design, as a general procedure, neural network generalised minimum variance self-tuning controller for non-linear dynamic plants (Integration of neural network mapping with generalised minimum variance self-tuning controller strategies); 3. To develop a software package to evaluate control system performances using Matlab, Simulink and Neural Network toolbox. An adaptive control algorithm utilising a recurrent network as a model of a partial unknown non-linear plant with unmeasurable state is proposed. Appropriately, it appears that structured recurrent neural networks can provide conveniently parameterised dynamic models for many non-linear systems for use in adaptive control. Properties of static neural networks, which enabled successful design of stable adaptive control in the state feedback case, are also identified. A survey of the existing results is presented which puts them in a systematic framework showing their relation to classical self-tuning adaptive control application of neural control to a SISO/MIMO control. Simulation results demonstrate that the self-tuning design methods may be practically applicable to a reasonably large class of unknown linear and non-linear dynamic control systems.

The development of a computer aided system for adaptive budgeting in a confectionery manufacturing firm

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY AND INFORMATION SERVICES WITH PRIOR ARRANGEMENT

Management information:design of a system for change identification

This thesis deals with the problem of Information Systems design for Corporate Management. It shows that the results of applying current approaches to Management Information Systems and Corporate Modelling fully justify a fresh look to the problem. The thesis develops an approach to design based on Cybernetic principles and theories. It looks at Management as an informational process and discusses the relevance of regulation theory to its practice. The work proceeds around the concept of change and its effects on the organization's stability and survival. The idea of looking at organizations as viable systems is discussed and a design to enhance survival capacity is developed. It takes Ashby's theory of adaptation and developments on ultra-stability as a theoretical framework and considering conditions for learning and foresight deduces that a design should include three basic components: A dynamic model of the organization- environment relationships; a method to spot significant changes in the value of the essential variables and in a certain set of parameters; and a Controller able to conceive and change the other two elements and to make choices among alternative policies. Further considerations of the conditions for rapid adaptation in organisms composed of many parts, and the law of Requisite Variety determine that successful adaptive behaviour requires certain functional organization. Beer's model of viable organizations is put in relation to Ashby's theory of adaptation and regulation. The use of the Ultra-stable system as abstract unit of analysis permits developing a rigorous taxonomy of change; it starts distinguishing between change with in behaviour and change of behaviour to complete the classification with organizational change. It relates these changes to the logical categories of learning connecting the topic of Information System design with that of organizational learning.

The development of calibration and adaptive filtering procedures for neuromagnetic studies of human visual function

This thesis first considers the calibration and signal processing requirements of a neuromagnetometer for the measurement of human visual function. Gradiometer calibration using straight wire grids is examined and optimal grid configurations determined, given realistic constructional tolerances. Simulations show that for gradiometer balance of 1:104 and wire spacing error of 0.25mm the achievable calibration accuracy of gain is 0.3%, of position is 0.3mm and of orientation is 0.6°. Practical results with a 19-channel 2nd-order gradiometer based system exceed this performance. The real-time application of adaptive reference noise cancellation filtering to running-average evoked response data is examined. In the steady state, the filter can be assumed to be driven by a non-stationary step input arising at epoch boundaries. Based on empirical measures of this driving step an optimal progression for the filter time constant is proposed which improves upon fixed time constant filter performance. The incorporation of the time-derivatives of the reference channels was found to improve the performance of the adaptive filtering algorithm by 15-20% for unaveraged data, falling to 5% with averaging. The thesis concludes with a neuromagnetic investigation of evoked cortical responses to chromatic and luminance grating stimuli. The global magnetic field power of evoked responses to the onset of sinusoidal gratings was shown to have distinct chromatic and luminance sensitive components. Analysis of the results, using a single equivalent current dipole model, shows that these components arise from activity within two distinct cortical locations. Co-registration of the resulting current source localisations with MRI shows a chromatically responsive area lying along the midline within the calcarine fissure, possibly extending onto the lingual and cuneal gyri. It is postulated that this area is the human homologue of the primate cortical area V4.

Adaptive electrical signal post-processing in optical communications systems

Improving bit error rates in optical communication systems is a difficult and important problem. The error correction must take place at high speed and be extremely accurate. We show the feasibility of using hardware implementable machine learning techniques. This may enable some error correction at the speed required.

Finding structures in information networks using the affinity network

This thesis proposes a novel graphical model for inference called the Affinity Network,which displays the closeness between pairs of variables and is an alternative to Bayesian Networks and Dependency Networks. The Affinity Network shares some similarities with Bayesian Networks and Dependency Networks but avoids their heuristic and stochastic graph construction algorithms by using a message passing scheme. A comparison with the above two instances of graphical models is given for sparse discrete and continuous medical data and data taken from the UCI machine learning repository. The experimental study reveals that the Affinity Network graphs tend to be more accurate on the basis of an exhaustive search with the small datasets. Moreover, the graph construction algorithm is faster than the other two methods with huge datasets. The Affinity Network is also applied to data produced by a synchronised system. A detailed analysis and numerical investigation into this dynamical system is provided and it is shown that the Affinity Network can be used to characterise its emergent behaviour even in the presence of noise.

Full-field feed-forward equalizer with adaptive system optimization

We numerically investigate the combination of full-field detection and feed-forward equalizer (FFE) for adaptive chromatic dispersion compensation up to 2160 km in a 10 Gbit/s on-off keyed optical transmission system. The technique, with respect to earlier reports, incorporates several important implementation modules, including the algorithm for adaptive equalization of the gain imbalance between the two receiver chains, compensation of phase misalignment of the asymmetric Mach-Zehnder interferometer, and simplified implementation of field calculation. We also show that in addition to enabling fast adaptation and simplification of field calculation, full-field FFE exhibits enhanced tolerance to the sampling phase misalignment and reduced sampling rate when compared to the full-field implementation using a dispersive transmission line.