A Connection Between GRBF and MLP

Both multilayer perceptrons (MLP) and Generalized Radial Basis Functions (GRBF) have good approximation properties, theoretically and experimentally. Are they related? The main point of this paper is to show that for normalized inputs, multilayer perceptron networks are radial function networks (albeit with a non-standard radial function). This provides an interpretation of the weights w as centers t of the radial function network, and therefore as equivalent to templates. This insight may be useful for practical applications, including better initialization procedures for MLP. In the remainder of the paper, we discuss the relation between the radial functions that correspond to the sigmoid for normalized inputs and well-behaved radial basis functions, such as the Gaussian. In particular, we observe that the radial function associated with the sigmoid is an activation function that is good approximation to Gaussian basis functions for a range of values of the bias parameter. The implication is that a MLP network can always simulate a Gaussian GRBF network (with the same number of units but less parameters); the converse is true only for certain values of the bias parameter. Numerical experiments indicate that this constraint is not always satisfied in practice by MLP networks trained with backpropagation. Multiscale GRBF networks, on the other hand, can approximate MLP networks with a similar number of parameters.

Millimeter-wave compressive holography.

We describe an active millimeter-wave holographic imaging system that uses compressive measurements for three-dimensional (3D) tomographic object estimation. Our system records a two-dimensional (2D) digitized Gabor hologram by translating a single pixel incoherent receiver. Two approaches for compressive measurement are undertaken: nonlinear inversion of a 2D Gabor hologram for 3D object estimation and nonlinear inversion of a randomly subsampled Gabor hologram for 3D object estimation. The object estimation algorithm minimizes a convex quadratic problem using total variation (TV) regularization for 3D object estimation. We compare object reconstructions using linear backpropagation and TV minimization, and we present simulated and experimental reconstructions from both compressive measurement strategies. In contrast with backpropagation, which estimates the 3D electromagnetic field, TV minimization estimates the 3D object that produces the field. Despite undersampling, range resolution is consistent with the extent of the 3D object band volume.

Second-Order Training of Adaptive Critics for Online Process Control

A new algorithm for training of nonlinear optimal neuro-controllers (in the form of the model-free, action-dependent, adaptive critic paradigm). Overcomes problems with existing stochastic backpropagation training: need for data storage, parameter shadowing and poor convergence, offering significant benefits for online applications.

Neural network forecasts of input-output technology

A significant part of the literature on input-output (IO) analysis is dedicated to the development and application of methodologies forecasting and updating technology coefficients and multipliers. Prominent among such techniques is the RAS method, while more information demanding econometric methods, as well as other less promising ones, have been proposed. However, there has been little interest expressed in the use of more modern and often more innovative methods, such as neural networks in IO analysis in general. This study constructs, proposes and applies a Backpropagation Neural Network (BPN) with the purpose of forecasting IO technology coefficients and subsequently multipliers. The RAS method is also applied on the same set of UK IO tables, and the discussion of results of both methods is accompanied by a comparative analysis. The results show that the BPN offers a valid alternative way of IO technology forecasting and many forecasts were more accurate using this method. Overall, however, the RAS method outperformed the BPN but the difference is rather small to be systematic and there are further ways to improve the performance of the BPN.

Predicting market movement with Artificial Neural Networks

This paper presents an application of an Artificial Neural Network (ANN) to the prediction of stock market direction in the US. Using a multilayer perceptron neural network and a backpropagation algorithm for the training process, the model aims at learning the hidden patterns in the daily movement of the S&P500 to correctly identify if the market will be in a Trend Following or Mean Reversion behavior. The ANN is able to produce a successful investment strategy which outperforms the buy and hold strategy, but presents instability in its overall results which compromises its practical application in real life investment decisions.

Algorithmes d’apprentissage profonds supervisés et non-supervisés: applications et résultats théoriques

La liste des domaines touchés par l’apprentissage machine s’allonge rapidement. Au fur et à mesure que la quantité de données disponibles augmente, le développement d’algorithmes d’apprentissage de plus en plus puissants est crucial. Ce mémoire est constitué de trois parties: d’abord un survol des concepts de bases de l’apprentissage automatique et les détails nécessaires pour l’entraînement de réseaux de neurones, modèles qui se livrent bien à des architectures profondes. Ensuite, le premier article présente une application de l’apprentissage machine aux jeux vidéos, puis une méthode de mesure performance pour ceux-ci en tant que politique de décision. Finalement, le deuxième article présente des résultats théoriques concernant l’entraînement d’architectures profondes nonsupervisées. Les jeux vidéos sont un domaine particulièrement fertile pour l’apprentissage automatique: il estf facile d’accumuler d’importantes quantités de données, et les applications ne manquent pas. La formation d’équipes selon un critère donné est une tˆache commune pour les jeux en lignes. Le premier article compare différents algorithmes d’apprentissage à des réseaux de neurones profonds appliqués à la prédiction de la balance d’un match. Ensuite nous présentons une méthode par simulation pour évaluer les modèles ainsi obtenus utilisés dans le cadre d’une politique de décision en ligne. Dans un deuxième temps nous présentons une nouvelleméthode pour entraîner des modèles génératifs. Des résultats théoriques nous indiquent qu’il est possible d’entraîner par rétropropagation des modèles non-supervisés pouvant générer des échantillons qui suivent la distribution des données. Ceci est un résultat pertinent dans le cadre de la récente littérature scientifique investiguant les propriétés des autoencodeurs comme modèles génératifs. Ces résultats sont supportés avec des expériences qualitatives préliminaires ainsi que quelques résultats quantitatifs.

AN ARTIFICIAL NEURAL NETWORK FOR SONAR TARGET DETECTION

Neural Network has emerged as the topic of the day. The spectrum of its application is as wide as from ECG noise filtering to seismic data analysis and from elementary particle detection to electronic music composition. The focal point of the proposed work is an application of a massively parallel connectionist model network for detection of a sonar target. This task is segmented into: (i) generation of training patterns from sea noise that contains radiated noise of a target, for teaching the network;(ii) selection of suitable network topology and learning algorithm and (iii) training of the network and its subsequent testing where the network detects, in unknown patterns applied to it, the presence of the features it has already learned in. A three-layer perceptron using backpropagation learning is initially subjected to a recursive training with example patterns (derived from sea ambient noise with and without the radiated noise of a target). On every presentation, the error in the output of the network is propagated back and the weights and the bias associated with each neuron in the network are modified in proportion to this error measure. During this iterative process, the network converges and extracts the target features which get encoded into its generalized weights and biases.In every unknown pattern that the converged network subsequently confronts with, it searches for the features already learned and outputs an indication for their presence or absence. This capability for target detection is exhibited by the response of the network to various test patterns presented to it.Three network topologies are tried with two variants of backpropagation learning and a grading of the performance of each combination is subsequently made.

Development of a Biometric Personal Authentication System Based on Fingerprint and Speech

Biometrics deals with the physiological and behavioral characteristics of an individual to establish identity. Fingerprint based authentication is the most advanced biometric authentication technology. The minutiae based fingerprint identification method offer reasonable identification rate. The feature minutiae map consists of about 70-100 minutia points and matching accuracy is dropping down while the size of database is growing up. Hence it is inevitable to make the size of the fingerprint feature code to be as smaller as possible so that identification may be much easier. In this research, a novel global singularity based fingerprint representation is proposed. Fingerprint baseline, which is the line between distal and intermediate phalangeal joint line in the fingerprint, is taken as the reference line. A polygon is formed with the singularities and the fingerprint baseline. The feature vectors are the polygonal angle, sides, area, type and the ridge counts in between the singularities. 100% recognition rate is achieved in this method. The method is compared with the conventional minutiae based recognition method in terms of computation time, receiver operator characteristics (ROC) and the feature vector length. Speech is a behavioural biometric modality and can be used for identification of a speaker. In this work, MFCC of text dependant speeches are computed and clustered using k-means algorithm. A backpropagation based Artificial Neural Network is trained to identify the clustered speech code. The performance of the neural network classifier is compared with the VQ based Euclidean minimum classifier. Biometric systems that use a single modality are usually affected by problems like noisy sensor data, non-universality and/or lack of distinctiveness of the biometric trait, unacceptable error rates, and spoof attacks. Multifinger feature level fusion based fingerprint recognition is developed and the performances are measured in terms of the ROC curve. Score level fusion of fingerprint and speech based recognition system is done and 100% accuracy is achieved for a considerable range of matching threshold

Optimal Unsupervised Learning in Feedforward Neural Networks

We investigate the properties of feedforward neural networks trained with Hebbian learning algorithms. A new unsupervised algorithm is proposed which produces statistically uncorrelated outputs. The algorithm causes the weights of the network to converge to the eigenvectors of the input correlation with largest eigenvalues. The algorithm is closely related to the technique of Self-supervised Backpropagation, as well as other algorithms for unsupervised learning. Applications of the algorithm to texture processing, image coding, and stereo depth edge detection are given. We show that the algorithm can lead to the development of filters qualitatively similar to those found in primate visual cortex.

Comparing Support Vector Machines with Gaussian Kernels to Radial Basis Function Classifiers

The Support Vector (SV) machine is a novel type of learning machine, based on statistical learning theory, which contains polynomial classifiers, neural networks, and radial basis function (RBF) networks as special cases. In the RBF case, the SV algorithm automatically determines centers, weights and threshold such as to minimize an upper bound on the expected test error. The present study is devoted to an experimental comparison of these machines with a classical approach, where the centers are determined by $k$--means clustering and the weights are found using error backpropagation. We consider three machines, namely a classical RBF machine, an SV machine with Gaussian kernel, and a hybrid system with the centers determined by the SV method and the weights trained by error backpropagation. Our results show that on the US postal service database of handwritten digits, the SV machine achieves the highest test accuracy, followed by the hybrid approach. The SV approach is thus not only theoretically well--founded, but also superior in a practical application.

Application of neural network to hybrid systems with binary inputs

Boolean input systems are in common used in the electric industry. Power supplies include such systems and the power converter represents these. For instance, in power electronics, the control variable are the switching ON and OFF of components as thyristors or transistors. The purpose of this paper is to use neural network (NN) to control continuous systems with Boolean inputs. This method is based on classification of system variations associated with input configurations. The classical supervised backpropagation algorithm is used to train the networks. The training of the artificial neural network and the control of Boolean input systems are presented. The design procedure of control systems is implemented on a nonlinear system. We apply those results to control an electrical system composed of an induction machine and its power converter.

Neurofuzzy network model construction using Bézier-Bernstein polynomial functions

Neurofuzzy modelling systems combine fuzzy logic with quantitative artificial neural networks via a concept of fuzzification by using a fuzzy membership function usually based on B-splines and algebraic operators for inference, etc. The paper introduces a neurofuzzy model construction algorithm using Bezier-Bernstein polynomial functions as basis functions. The new network maintains most of the properties of the B-spline expansion based neurofuzzy system, such as the non-negativity of the basis functions, and unity of support but with the additional advantages of structural parsimony and Delaunay input space partitioning, avoiding the inherent computational problems of lattice networks. This new modelling network is based on the idea that an input vector can be mapped into barycentric co-ordinates with respect to a set of predetermined knots as vertices of a polygon (a set of tiled Delaunay triangles) over the input space. The network is expressed as the Bezier-Bernstein polynomial function of barycentric co-ordinates of the input vector. An inverse de Casteljau procedure using backpropagation is developed to obtain the input vector's barycentric co-ordinates that form the basis functions. Extension of the Bezier-Bernstein neurofuzzy algorithm to n-dimensional inputs is discussed followed by numerical examples to demonstrate the effectiveness of this new data based modelling approach.

Neural network based control in food processing

A number of commonly encountered simple neural network types are discussed, with particular attention being paid to their applicability in automation and control when applied to food processing. In the first instance n-tuple networks are considered, these being particularly useful for high speed production checking operations. Subsequently backpropagation networks are discussed, these being useful both in a more familiar feedback control arrangement and also for such things as recipe prediction.

A model of UK domestic appliance-use using Artificial Neural Networks

The development of an Artificial Neural Network model of UK domestic appliance energy consumption is presented. The model uses diary-style appliance use data and a survey questionnaire collected from 51 households during the summer of 2010. It also incorporates measured energy data and is sensitive to socioeconomic, physical dwelling and temperature variables. A prototype model is constructed in MATLAB using a two layer feed forward network with backpropagation training and has a12:10:24architecture.Model outputs include appliance load profiles which can be applied to the fields of energy planning (micro renewables and smart grids), building simulation tools and energy policy.

Avaliação do desempenho de classificadores neurais para aplicações em sensoriamento remoto

Atualmente, pesquisadores das mais diversas áreas, tais como: Geologia, Física, Cartografia, Oceanografia, entre outras, utilizam imagens de satélite como uma fonte valiosa para a extração de informações sobre a superfície terrestre. Muitas vezes, a análise (classificação) destas imagens é realizada por métodos tradicionais sejam eles supervisionados (como o Método de Máxima Verossimilhança Gaussiana) ou nãosupervisionados (como o Método de Seleção pelo Pico do Histograma). Entretanto, pode-se utilizar as Redes Neurais Artificiais como uma alternativa para o aumento da acurácia em classificações digitais. Neste trabalho, utilizou-se imagens multi-espectrais do satélite LANDSAT 5-TM para a identificação de espécies vegetais (Mata Nativa, Eucalyptus e Acácia) em uma região próxima aos municípios de General Câmara, Santo Amaro e Taquari, no Estado do Rio Grande do Sul, Brasil. Comparou-se qualitativamente e quantitativamente os resultados obtidos pelo método de Máxima Verossimilhança Gaussiana e por uma Rede Neural Artificial Multinível com BackPropagation na classificação da área de estudo. Para tanto, parte desta área foi mapeada através de uma verificação de campo e com o auxílio de classificadores nãosupervisionados (Kohonen, que é uma Rede Neural, e o método de Seleção pelo Pico do Histograma). Com isto, foi possível coletar dois conjuntos de amostras, sendo que um deles foi utilizado para o treinamento dos métodos e o outro (conjunto de reconhecimento) serviu para a avaliação das classificações obtidas. Após o treinamento, parte da área de estudo foi classificada por ambos os métodos. Em seguida, os resultados obtidos foram avaliados através do uso de Tabelas de Contingência, considerando um nível de significância de 5%. Por fim, na maior parte dos testes realizados, a Rede Neural Artificial Multinível com BackPropagation apresentou valores de acurácia superiores ao Método de Máxima Verossimilhança Gaussiana. Assim, com este trabalho observou-se que não há diferença significativa de classificação para as espécies vegetais, ao nível de 5%, para a área de estudo considerada, na época de aquisição da imagem, para o conjunto de reconhecimento.