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.

A novel radial basis function neural network principal component analysis scheme for PMU-based wide-area power system monitoring

A novel model-based principal component analysis (PCA) method is proposed in this paper for wide-area power system monitoring, aiming to tackle one of the critical drawbacks of the conventional PCA, i.e. the incapability to handle non-Gaussian distributed variables. It is a significant extension of the original PCA method which has already shown to outperform traditional methods like rate-of-change-of-frequency (ROCOF). The ROCOF method is quick for processing local information, but its threshold is difficult to determine and nuisance tripping may easily occur. The proposed model-based PCA method uses a radial basis function neural network (RBFNN) model to handle the nonlinearity in the data set to solve the no-Gaussian issue, before the PCA method is used for islanding detection. To build an effective RBFNN model, this paper first uses a fast input selection method to remove insignificant neural inputs. Next, a heuristic optimization technique namely Teaching-Learning-Based-Optimization (TLBO) is adopted to tune the nonlinear parameters in the RBF neurons to build the optimized model. The novel RBFNN based PCA monitoring scheme is then employed for wide-area monitoring using the residuals between the model outputs and the real PMU measurements. Experimental results confirm the efficiency and effectiveness of the proposed method in monitoring a suite of process variables with different distribution characteristics, showing that the proposed RBFNN PCA method is a reliable scheme as an effective extension to the linear PCA method.

Artificial Neural Network for Compositional Ionic Liquid Viscosity Prediction

Being a new generation of green solvents and high-tech reaction media of the future, ionic liquids have increasingly attracted much attention. Of particular interest in this context are room temperature ionic liquids (in short as ILs in this paper). Due to the relatively high viscosity, ILs is expected to be used in the form of solvent diluted mixture with reduced viscosity in industrial application, where predicting the viscosity of IL mixture has been an important research issue. Different IL mixture and many modelling approaches have been investigated. The objective of this study is to provide an alternative model approach using soft computing technique, i.e., artificial neural network (ANN) model, to predict the compositional viscosity of binary mixtures of ILs [C _n-mim][NTf ₂] with n=4, 6, 8, 10 in methanol and ethanol over the entire range of molar fraction at a broad range of temperatures from T=293.0-328.0K. The results show that the proposed ANN model provides alternative way to predict compositional viscosity successfully with highly improved accuracy and also show its potential to be extensively utilized to predict compositional viscosity taking account of IL alkyl chain length, as well as temperature and compositions simultaneously, i.e., more complex intermolecular interactions between components in which it would be hard or impossible to establish the analytical model. This illustrates the potential application of ANN in the case that the physical and thermodynamic properties are highly non-linear or too complex. © 2012 Copyright the authors.

Artificial neural network model to predict compositional viscosity over a broad range of temperatures

The objective of this study is to provide an alternative model approach, i.e., artificial neural network (ANN) model, to predict the compositional viscosity of binary mixtures of room temperature ionic liquids (in short as ILs) [C _n-mim] [NTf ₂] with n=4, 6, 8, 10 in methanol and ethanol over the entire range of molar fraction at a broad range of temperatures from T=293.0328.0K. The results show that the proposed ANN model provides alternative way to predict compositional viscosity successfully with highly improved accuracy and also show its potential to be extensively utilized to predict compositional viscosity over a wide range of temperatures and more complex viscosity compositions, i.e., more complex intermolecular interactions between components in which it would be hard or impossible to establish the analytical model. © 2010 IEEE.

Neural Network Based Attack on a Masked Implementation of AES

Masked implementations of cryptographic algorithms are often used in commercial embedded cryptographic devices to increase their resistance to side channel attacks. In this work we show how neural networks can be used to both identify the mask value, and to subsequently identify the secret key value with a single attack trace with high probability. We propose the use of a pre-processing step using principal component analysis (PCA) to significantly increase the success of the attack. We have developed a classifier that can correctly identify the mask for each trace, hence removing the security provided by that mask and reducing the attack to being equivalent to an attack against an unprotected implementation. The attack is performed on the freely available differential power analysis (DPA) contest data set to allow our work to be easily reproducible. We show that neural networks allow for a robust and efficient classification in the context of side-channel attacks.

A Spiking Neural Network Model of the Medial Superior Olive using Spike Timing Dependent Plasticity for Sound Localisation

Sound localization can be defined as the ability to identify the position of an input sound source and is considered a powerful aspect of mammalian perception. For low frequency sounds, i.e., in the range 270 Hz-1.5 KHz, the mammalian auditory pathway achieves this by extracting the Interaural Time Difference between sound signals being received by the left and right ear. This processing is performed in a region of the brain known as the Medial Superior Olive (MSO). This paper presents a Spiking Neural Network (SNN) based model of the MSO. The network model is trained using the Spike Timing Dependent Plasticity learning rule using experimentally observed Head Related Transfer Function data in an adult domestic cat. The results presented demonstrate how the proposed SNN model is able to perform sound localization with an accuracy of 91.82% when an error tolerance of +/-10 degrees is used. For angular resolutions down to 2.5 degrees , it will be demonstrated how software based simulations of the model incur significant computation times. The paper thus also addresses preliminary implementation on a Field Programmable Gate Array based hardware platform to accelerate system performance.

Spiking neural network model of sound localisation using the interaural intensity difference

In this paper, a spiking neural network (SNN) architecture to simulate the sound localization ability of the mammalian auditory pathways using the interaural intensity difference cue is presented. The lateral superior olive was the inspiration for the architecture, which required the integration of an auditory periphery (cochlea) model and a model of the medial nucleus of the trapezoid body. The SNN uses leaky integrateand-fire excitatory and inhibitory spiking neurons, facilitating synapses and receptive fields. Experimentally derived headrelated transfer function (HRTF) acoustical data from adult domestic cats were employed to train and validate the localization ability of the architecture, training used the supervised learning algorithm called the remote supervision method to determine the azimuthal angles. The experimental results demonstrate that the architecture performs best when it is localizing high-frequency sound data in agreement with the biology, and also shows a high degree of robustness when the HRTF acoustical data is corrupted by noise.