Improving neural network training solutions using regularisation

This paper describes the application of regularisation to the training of feedforward neural networks, as a means of improving the quality of solutions obtained. The basic principles of regularisation theory are outlined for both linear and nonlinear training and then extended to cover a new hybrid training algorithm for feedforward neural networks recently proposed by the authors. The concept of functional regularisation is also introduced and discussed in relation to MLP and RBF networks. The tendency for the hybrid training algorithm and many linear optimisation strategies to generate large magnitude weight solutions when applied to ill-conditioned neural paradigms is illustrated graphically and reasoned analytically. While such weight solutions do not generally result in poor fits, it is argued that they could be subject to numerical instability and are therefore undesirable. Using an illustrative example it is shown that, as well as being beneficial from a generalisation perspective, regularisation also provides a means for controlling the magnitude of solutions. (C) 2001 Elsevier Science B.V. All rights reserved.

Artificial Intelligence Model for Flowable Concrete Mixtures Used in Underwater Construction and Repair

This study explores using artificial neural networks to predict the rheological and mechanical properties of underwater concrete (UWC) mixtures and to evaluate the sensitivity of such properties to variations in mixture ingredients. Artificial neural networks (ANN) mimic the structure and operation of biological neurons and have the unique ability of self-learning, mapping, and functional approximation. Details of the development of the proposed neural network model, its architecture, training, and validation are presented in this study. A database incorporating 175 UWC mixtures from nine different studies was developed to train and test the ANN model. The data are arranged in a patterned format. Each pattern contains an input vector that includes quantity values of the mixture variables influencing the behavior of UWC mixtures (that is, cement, silica fume, fly ash, slag, water, coarse and fine aggregates, and chemical admixtures) and a corresponding output vector that includes the rheological or mechanical property to be modeled. Results show that the ANN model thus developed is not only capable of accurately predicting the slump, slump-flow, washout resistance, and compressive strength of underwater concrete mixtures used in the training process, but it can also effectively predict the aforementioned properties for new mixtures designed within the practical range of the input parameters used in the training process with an absolute error of 4.6, 10.6, 10.6, and 4.4%, respectively.

A fast multi-output RBF neural network construction method

This paper investigates the center selection of multi-output radial basis function (RBF) networks, and a multi-output fast recursive algorithm (MFRA) is proposed. This method can not only reveal the significance of each candidate center based on the reduction in the trace of the error covariance matrix, but also can estimate the network weights simultaneously using a back substitution approach. The main contribution is that the center selection procedure and the weight estimation are performed within a well-defined regression context, leading to a significantly reduced computational complexity. The efficiency of the algorithm is confirmed by a computational complexity analysis, and simulation results demonstrate its effectiveness. (C) 2010 Elsevier B.V. All rights reserved.

ARE ARTIFICIAL NEURAL NETS AS SMART AS A RAT

The origins of artificial neural networks are related to animal conditioning theory: both are forms of connectionist theory, which in turn derives from the empiricist philosophers' principle of association. The parallel between animal learning and neural nets suggests that interaction between them should benefit both sides.

Two-Stage Orthogonal Least Squares Methods for Neural Network Construction

A number of neural networks can be formulated as the linear-in-the-parameters models. Training such networks can be transformed to a model selection problem where a compact model is selected from all the candidates using subset selection algorithms. Forward selection methods are popular fast subset selection approaches. However, they may only produce suboptimal models and can be trapped into a local minimum. More recently, a two-stage fast recursive algorithm (TSFRA) combining forward selection and backward model refinement has been proposed to improve the compactness and generalization performance of the model. This paper proposes unified two-stage orthogonal least squares methods instead of the fast recursive-based methods. In contrast to the TSFRA, this paper derives a new simplified relationship between the forward and the backward stages to avoid repetitive computations using the inherent orthogonal properties of the least squares methods. Furthermore, a new term exchanging scheme for backward model refinement is introduced to reduce computational demand. Finally, given the error reduction ratio criterion, effective and efficient forward and backward subset selection procedures are proposed. Extensive examples are presented to demonstrate the improved model compactness constructed by the proposed technique in comparison with some popular methods.

Spiking neural network connectivity and its potential for temporal sensory processing and variable binding

The most biologically-inspired artificial neurons are those of the third generation, and are termed spiking neurons, as individual pulses or spikes are the means by which stimuli are communicated. In essence, a spike is a short-term change in electrical potential and is the basis of communication between biological neurons. Unlike previous generations of artificial neurons, spiking neurons operate in the temporal domain, and exploit time as a resource in their computation. In 1952, Alan Lloyd Hodgkin and Andrew Huxley produced the first model of a spiking neuron; their model describes the complex electro-chemical process that enables spikes to propagate through, and hence be communicated by, spiking neurons. Since this time, improvements in experimental procedures in neurobiology, particularly with in vivo experiments, have provided an increasingly more complex understanding of biological neurons. For example, it is now well-understood that the propagation of spikes between neurons requires neurotransmitter, which is typically of limited supply. When the supply is exhausted neurons become unresponsive. The morphology of neurons, number of receptor sites, amongst many other factors, means that neurons consume the supply of neurotransmitter at different rates. This in turn produces variations over time in the responsiveness of neurons, yielding various computational capabilities. Such improvements in the understanding of the biological neuron have culminated in a wide range of different neuron models, ranging from the computationally efficient to the biologically realistic. These models enable the modeling of neural circuits found in the brain.

Spiking Neural Network Connectivity and its Potential for Temporal Sensory Processing and Variable Binding

The most biologically-inspired artificial neurons are those of the third generation, and are termed spiking neurons, as individual pulses or spikes are the means by which stimuli are communicated. In essence, a spike is a short-term change in electrical potential and is the basis of communication between biological neurons. Unlike previous generations of artificial neurons, spiking neurons operate in the temporal domain, and exploit time as a resource in their computation. In 1952, Alan Lloyd Hodgkin and Andrew Huxley produced the first model of a spiking neuron; their model describes the complex electro-chemical process that enables spikes to propagate through, and hence be communicated by, spiking neurons. Since this time, improvements in experimental procedures in neurobiology, particularly with in vivo experiments, have provided an increasingly more complex understanding of biological neurons. For example, it is now well understood that the propagation of spikes between neurons requires neurotransmitter, which is typically of limited supply. When the supply is exhausted neurons become unresponsive. The morphology of neurons, number of receptor sites, amongst many other factors, means that neurons consume the supply of neurotransmitter at different rates. This in turn produces variations over time in the responsiveness of neurons, yielding various computational capabilities. Such improvements in the understanding of the biological neuron have culminated in a wide range of different neuron models, ranging from the computationally efficient to the biologically realistic. These models enable the modelling of neural circuits found in the brain. In recent years, much of the focus in neuron modelling has moved to the study of the connectivity of spiking neural networks. Spiking neural networks provide a vehicle to understand from a computational perspective, aspects of the brain’s neural circuitry. This understanding can then be used to tackle some of the historically intractable issues with artificial neurons, such as scalability and lack of variable binding. Current knowledge of feed-forward, lateral, and recurrent connectivity of spiking neurons, and the interplay between excitatory and inhibitory neurons is beginning to shed light on these issues, by improved understanding of the temporal processing capabilities and synchronous behaviour of biological neurons. This research topic aims to amalgamate current research aimed at tackling these phenomena.

ANN based support for distributed energy resources scheduling in smart grids

The future scenarios for operation of smart grids are likely to include a large diversity of players, of different types and sizes. With control and decision making being decentralized over the network, intelligence should also be decentralized so that every player is able to play in the market environment. In the new context, aggregator players, enabling medium, small, and even micro size players to act in a competitive environment, will be very relevant. Virtual Power Players (VPP) and single players must optimize their energy resource management in order to accomplish their goals. This is relatively easy to larger players, with financial means to have access to adequate decision support tools, to support decision making concerning their optimal resource schedule. However, the smaller players have difficulties in accessing this kind of tools. So, it is required that these smaller players can be offered alternative methods to support their decisions. This paper presents a methodology, based on Artificial Neural Networks (ANN), intended to support smaller players’ resource scheduling. The used methodology uses a training set that is built using the energy resource scheduling solutions obtained with a reference optimization methodology, a mixed-integer non-linear programming (MINLP) in this case. The trained network is able to achieve good schedule results requiring modest computational means.

Neural Network based studies on Spectroscopic Analysis and Image Processing

International School of Photonics, Cochin University of Science and Technology

Learning disability prediction tool using ANN and ANFIS

Learning Disability (LD) is a neurological condition that affects a child’s brain and impairs his ability to carry out one or many specific tasks. LD affects about 15 % of children enrolled in schools. The prediction of LD is a vital and intricate job. The aim of this paper is to design an effective and powerful tool, using the two intelligent methods viz., Artificial Neural Network and Adaptive Neuro-Fuzzy Inference System, for measuring the percentage of LD that affected in school-age children. In this study, we are proposing some soft computing methods in data preprocessing for improving the accuracy of the tool as well as the classifier. The data preprocessing is performed through Principal Component Analysis for attribute reduction and closest fit algorithm is used for imputing missing values. The main idea in developing the LD prediction tool is not only to predict the LD present in children but also to measure its percentage along with its class like low or minor or major. The system is implemented in Mathworks Software MatLab 7.10. The results obtained from this study have illustrated that the designed prediction system or tool is capable of measuring the LD effectively

Implementation of a Neural Network Classifier for Noise Sources in the Ocean

The paper investigates the feasibility of implementing an intelligent classifier for noise sources in the ocean, with the help of artificial neural networks, using higher order spectral features. Non-linear interactions between the component frequencies of the noise data can give rise to certain phase relations called Quadratic Phase Coupling (QPC), which cannot be characterized by power spectral analysis. However, bispectral analysis, which is a higher order estimation technique, can reveal the presence of such phase couplings and provide a measure to quantify such couplings. A feed forward neural network has been trained and validated with higher order spectral features

Dealing with complexity: an overview of the neural network approach

The problem of complexity is particularly relevant to the field of control engineering, since many engineering problems are inherently complex. The inherent complexity is such that straightforward computational problem solutions often produce very poor results. Although parallel processing can alleviate the problem to some extent, it is artificial neural networks (in various forms) which have recently proved particularly effective, even in dealing with the causes of the problem itself. This paper presents an overview of the current neural network research being undertaken. Such research aims to solve the complex problems found in many areas of science and engineering today.

Identification of nonlinear systems with a dynamic recurrent neural network

Two approaches are presented to calculate the weights for a Dynamic Recurrent Neural Network (DRNN) in order to identify the input-output dynamics of a class of nonlinear systems. The number of states of the identified network is constrained to be the same as the number of states of the plant.

Pricing and hedging short sterling options using artificial neural networks

This paper compares the performance of artificial neural networks (ANNs) with that of the modified Black model in both pricing and hedging Short Sterling options. Using high frequency data, standard and hybrid ANNs are trained to generate option prices. The hybrid ANN is significantly superior to both the modified Black model and the standard ANN in pricing call and put options. Hedge ratios for hedging Short Sterling options positions using Short Sterling futures are produced using the standard and hybrid ANN pricing models, the modified Black model, and also standard and hybrid ANNs trained directly on the hedge ratios. The performance of hedge ratios from ANNs directly trained on actual hedge ratios is significantly superior to those based on a pricing model, and to the modified Black model.