We use sampling methods to analyse the “apparent minima” of the error surfaces of feedforward neural networks learning encoder problems. First and second-order statistics of a sample of these points of attraction are shown to provide qualitative statistical information about the structure of the error surface, allowing a simple description of this structure. Following methods previously used in the analysis of other complex configuration spaces (such as spin glass models and several combinatorial optimization problems), the third-order statistics of the points of attraction are examined and found to be arranged in a highly ultrametric way, using the normal Euclidean distance measure. The implications of this result are discussed.

We present a constructive neural network training algorithm. The algorithm builds a single hidden layer network and is guaranteed to implement any consistent training set of real-valued vectors classified into any number of classes. The algorithm extends the ‘sequential’ learning algorithm of (Marchand et al., 1990) from the binary input case to the real valued input case.

In the last few years several alternatives to the mean-squared error function have been proposed as a means of improving the speed of backpropagation learning. Perhaps best-known among these are the functions considered in (Fahlman 1988) and the cross-entropy function discussed in (Solla et al., 1988). In this paper we investigate properties of error functions that influence the rate of learning when backpropagation (BP) is employed.

Foreign exchange trading has emerged in recent times as a significant activity in many countries. As with most forms of trading, the activity is influenced by many random parameters, so that the creation of a system that effectively emulates the trading process will be very helpful. This chapter presents a novel trading system using Machine Learning methods of Genetic Algorithms and Reinforcement Learning. The system emulates trader behavior on the Foreign Exchange market and finds the most profitable trading strategy.

The National Electricity Market in Australia became deregulated in 1998, providing competition at a retail level. As with most forms of trading, the activity is influenced by many random parameters so that the creation of a system that effectively emulates the trading process will be very helpful. A major issue for traders in the deregulated market is when to sell and when to buy electricity instruments in order to maximize profit. This paper presents novel trading strategies based on the Machine Learning methods of Genetic Algorithms and Reinforcement Learning for trading in electricity futures.

Foreign Exchange trading has emerged in recent times as a significant activity in many countries. As with most forms of trading, the activity is influenced by many random parameters so that the creation of a system that effectively emulates the trading process will be very helpful. In this paper we try to create such a system using Machine learning approach to emulate trader behaviour on the Foreign Exchange market and to find the most profitable trading strategy.

We present the results of applying automated machine learning techniques to the problem of matching different object catalogues in astrophysics. In this study, we take two partially matched catalogues where one of the two catalogues has a large positional uncertainty. The two catalogues we used here were taken from the H I Parkes All Sky Survey (HIPASS) and SuperCOSMOS optical survey. Previous work had matched 44 per cent (1887 objects) of HIPASS to the SuperCOSMOS catalogue.A supervised learning algorithm was then applied to construct a model of the matched portion of our catalogue. Validation of the model shows that we achieved a good classification performance (99.12 per cent correct).Applying this model to the unmatched portion of the catalogue found 1209 new matches. This increases the catalogue size from 1887 matched objects to 3096. The combination of these procedures yields a catalogue that is 72 per cent matched.

An emerging issue in the field of astronomy is the integration, management and utilization of databases from around the world to facilitate scientific discovery. In this paper, we investigate application of the machine learning techniques of support vector machines and neural networks to the problem of amalgamating catalogues of galaxies as objects from two disparate data sources: radio and optical. Formulating this as a classification problem presents several challenges, including dealing with a highly unbalanced data set. Unlike the conventional approach to the problem (which is based on a likelihood ratio) machine learning does not require density estimation and is shown here to provide a significant improvement in performance. We also report some experiments that explore the importance of the radio and optical data features for the matching problem.

The Tree Augmented Naïve Bayes (TAN) classifier relaxes the sweeping independence assumptions of the Naïve Bayes approach by taking account of conditional probabilities. It does this in a limited sense, by incorporating the conditional probability of each attribute given the class and (at most) one other attribute. The method of boosting has previously proven very effective in improving the performance of Naïve Bayes classifiers and in this paper, we investigate its effectiveness on application to the TAN classifier.

This paper presents the implementation of a modified particle filter for vision-based simultaneous localization and mapping of an autonomous robot in a structured indoor environment. Through this method, artificial landmarks such as multi-coloured cylinders can be tracked with a camera mounted on the robot, and the position of the robot can be estimated at the same time. Experimental results in simulation and in real environments show that this approach has advantages over the extended Kalman filter with ambiguous data association and various levels of odometric noise.

In this paper we demonstrate that it is possible to gradually improve the performance of support vector machine (SVM) classifiers by using a genetic algorithm to select a sequence of training subsets from the available data. Performance improvement is possible because the SVM solution generally lies some distance away from the Bayes optimal in the space of learning parameters. We illustrate performance improvements on a number of benchmark data sets.

Loss allocation of electricity has been in the limelight since the introduction of deregulation in the electricity market. For a fair and transparent market, method to allocate appropriate loss to any customer is necessary. This paper paper proposes an approximated loss allocation method based on network reduction. Reduction of the network is dependent upon bilateral contract between any one generator (seller)and one load (customer/buyer). Gaussian elemination is applied to reduce the bus admittance matrix of the network. The reduced network will contain only buses of primary interest - that is of the analysed contract. IEEE 14-bus test system is used to illustrate the proposed methodology. The results from this analysis are presented in this paper.

Support vector machines (SVMs) have recently emerged as a powerful technique for solving problems in pattern classification and regression. Best performance is obtained from the SVM its parameters have their values optimally set. In practice, good parameter settings are usually obtained by a lengthy process of trial and error. This paper describes the use of genetic algorithm to evolve these parameter settings for an application in mobile robotics.

Deregulation has been one of the most intriguing issues in the electricity market for the past decade. It has transformed many aspects of the market, including electricity pricing. Prior to deregulation, generators and transmission lines operated as one entity. Therefore it seemed reasonable to allocate approximately 2% to 5% of generation as power loss. In light of deregulation, it is important that usage allocation and loss allocation are determined so that the market operates fairly and transparently. The preliminary findings that will be discussed in this paper include deregulation in Australia, nodal pricing globally, existing allocation methods and some findings from initial investigations into the methods.

In this paper we present an analysis of the results of a study into wholesale (spot) electricity price forecasting with Support Vector Machines (SVM) utilising past price and demand data and Projected Assessment of System Adequacy (PASA) data. The forecasting accuracy was evaluated using Australian National Electricity Market (NEM), New South Wales regional data over the year 2002. The inclusion of PASA data shows little improvement in forecasting accuracy.

This paper gives a brief description of a hierarchical architecture (HONG) that has been described in Atukorale and Suganthan (Neurocomputing 35 (2000) 165). The learning algorithm it uses is a mixed unsupervised/supervised method with most of the learning being unsupervised. The architecture generates multiple classifications for every data pattern presented, and combines them to obtain the final classification. The main objective of this paper is to show how boosting can be used to improve the performance of the HONG classifier.

This paper is concerned with the use of scientific visualization methods for the analysis of feedforward neural networks (NNs). Inevitably, the kinds of data associated with the design and implementation of neural networks are of very high dimensionality, presenting a major challenge for visualization. A method is described using the well-known statistical technique of principal component analysis (PCA). This is found to be an effective and useful method of visualizing the learning trajectories of many learning algorithms such as back-propagation and can also be used to provide insight into the learning process and the nature of the error surface.

Foreign exchange trading has emerged in recent times as a significant activity in many countries. As with most forms of trading, the activity is influenced by many random parameters so that the creation of a system that effectively emulates the trading process is very helpful. In this paper, we try to create such a system with a genetic algorithm engine to emulate trader behaviour on the foreign exchange market and to find the most profitable trading strategy.