The Usage of Neural Networks for the Medical Diagnosis

The problem of cancer diagnosis from multi-channel images using the neural networks is investigated. The goal of this work is to classify the different tissue types which are used to determine the cancer risk. The radial basis function networks and backpropagation neural networks are used for classification. The results of experiments are presented.

Examining the use of bid information in predicting the contractor’s performance

Purpose – The purpose of this paper is to examine the use of bid information, including both price and non-price factors in predicting the bidder’s performance. Design/methodology/approach – The practice of the industry was first reviewed. Data on bid evaluation and performance records of the successful bids were then obtained from the Hong Kong Housing Department, the largest housing provider in Hong Kong. This was followed by the development of a radial basis function (RBF) neural network based performance prediction model. Findings – It is found that public clients are more conscientious and include non-price factors in their bid evaluation equations. With the input variables used the information is available at the time of the bid and the output variable is the project performance score recorded during work in progress achieved by the successful bidder. It was found that past project performance score is the most sensitive input variable in predicting future performance. Research limitations/implications – The paper shows the inadequacy of using price alone for bid award criterion. The need for a systemic performance evaluation is also highlighted, as this information is highly instrumental for subsequent bid evaluations. The caveat for this study is that the prediction model was developed based on data obtained from one single source. Originality/value – The value of the paper is in the use of an RBF neural network as the prediction tool because it can model non-linear function. This capability avoids tedious ‘‘trial and error’’ in deciding the number of hidden layers to be used in the network model. Keywords Hong Kong, Construction industry, Neural nets, Modelling, Bid offer spreads Paper type Research paper

A model predictive approach for community battery energy storage system optimization

This paper presents an efficient algorithm for optimizing the operation of battery storage in a low voltage distribution network with a high penetration of PV generation. A predictive control solution is presented that uses wavelet neural networks to predict the load and PV generation at hourly intervals for twelve hours into the future. The load and generation forecast, and the previous twelve hours of load and generation history, is used to assemble load profile. A diurnal charging profile can be compactly represented by a vector of Fourier coefficients allowing a direct search optimization algorithm to be applied. The optimal profile is updated hourly allowing the state of charge profile to respond to changing forecasts in load.

Automatic UAV forced landing site detection using machine learning

The commercialization of aerial image processing is highly dependent on the platforms such as UAVs (Unmanned Aerial Vehicles). However, the lack of an automated UAV forced landing site detection system has been identified as one of the main impediments to allow UAV flight over populated areas in civilian airspace. This article proposes a UAV forced landing site detection system that is based on machine learning approaches including the Gaussian Mixture Model and the Support Vector Machine. A range of learning parameters are analysed including the number of Guassian mixtures, support vector kernels including linear, radial basis function Kernel (RBF) and polynormial kernel (poly), and the order of RBF kernel and polynormial kernel. Moreover, a modified footprint operator is employed during feature extraction to better describe the geometric characteristics of the local area surrounding a pixel. The performance of the presented system is compared to a baseline UAV forced landing site detection system which uses edge features and an Artificial Neural Network (ANN) region type classifier. Experiments conducted on aerial image datasets captured over typical urban environments reveal improved landing site detection can be achieved with an SVM classifier with an RBF kernel using a combination of colour and texture features. Compared to the baseline system, the proposed system provides significant improvement in term of the chance to detect a safe landing area, and the performance is more stable than the baseline in the presence of changes to the UAV altitude.

Improved Kalman filter initialisation using neurofuzzy estimation

It is traditional to initialise Kalman filters and extended Kalman filters with estimates of the states calculated directly from the observed (raw) noisy inputs, but unfortunately their performance is extremely sensitive to state initialisation accuracy: good initial state estimates ensure fast convergence whereas poor estimates may give rise to slow convergence or even filter divergence. Divergence is generally due to excessive observation noise and leads to error magnitudes that quickly become unbounded (R.J. Fitzgerald, 1971). When a filter diverges, it must be re initialised but because the observations are extremely poor, re initialised states will have poor estimates. The paper proposes that if neurofuzzy estimators produce more accurate state estimates than those calculated from the observed noisy inputs (using the known state model), then neurofuzzy estimates can be used to initialise the states of Kalman and extended Kalman filters. Filters whose states have been initialised with neurofuzzy estimates should give improved performance by way of faster convergence when the filter is initialised, and when a filter is re started after divergence

A sharable instructable agent for information filtering

In this paper we present an information filtering agent called sharable instructable information filtering agent (SIIFA). It adopted the approach of sharable instructable agents. SIIFA provides comprehensible and flexible interaction to represent and filter the documents. The representation scheme in SIIFA is personalized. It, either fully or partly, can be shared among the users of the stream while not revealing their interests and can be easily edited. SIIFA is evaluated on the comp.ai.neural-nets Usent newsgroup documents and compared with the vector space method.

电阻率测井反演方法应用研究

Formation resistivity is one of the most important parameters to be evaluated in the evaluation of reservoir. In order to acquire the true value of virginal formation, various types of resistivity logging tools have been developed. However, with the increment of the proved reserves, the thickness of interest pay zone is becoming thinner and thinner, especially in the terrestrial deposit oilfield, so that electrical logging tools, limited by the contradictory requirements of resolution and investigation depth of this kinds of tools, can not provide the true value of the formation resistivity. Therefore, resitivity inversion techniques have been popular in the determination of true formation resistivity based on the improving logging data from new tools. In geophysical inverse problems, non-unique solution is inevitable due to the noisy data and deficient measurement information. I address this problem in my dissertation from three aspects, data acquisition, data processing/inversion and applications of the results/ uncertainty evaluation of the non-unique solution. Some other problems in the traditional inversion methods such as slowness speed of the convergence and the initial-correlation results. Firstly, I deal with the uncertainties in the data to be processed. The combination of micro-spherically focused log (MSFL) and dual laterolog(DLL) is the standard program to determine formation resistivity. During the inversion, the readings of MSFL are regarded as the resistivity of invasion zone of the formation after being corrected. However, the errors can be as large as 30 percent due to mud cake influence even if the rugose borehole effects on the readings of MSFL can be ignored. Furthermore, there still are argues about whether the two logs can be quantitatively used to determine formation resisitivities due to the different measurement principles. Thus, anew type of laterolog tool is designed theoretically. The new tool can provide three curves with different investigation depths and the nearly same resolution. The resolution is about 0.4meter. Secondly, because the popular iterative inversion method based on the least-square estimation can not solve problems more than two parameters simultaneously and the new laterolog logging tool is not applied to practice, my work is focused on two parameters inversion (radius of the invasion and the resistivty of virgin information ) of traditional dual laterolog logging data. An unequal weighted damp factors- revised method is developed to instead of the parameter-revised techniques used in the traditional inversion method. In this new method, the parameter is revised not only dependency on the damp its self but also dependency on the difference between the measurement data and the fitting data in different layers. At least 2 iterative numbers are reduced than the older method, the computation cost of inversion is reduced. The damp least-squares inversion method is the realization of Tikhonov's tradeoff theory on the smooth solution and stability of inversion process. This method is realized through linearity of non-linear inversion problem which must lead to the dependency of solution on the initial value of parameters. Thus, severe debates on efficiency of this kinds of methods are getting popular with the developments of non-linear processing methods. The artificial neural net method is proposed in this dissertation. The database of tool's response to formation parameters is built through the modeling of the laterolog tool and then is used to training the neural nets. A unit model is put forward to simplify the dada space and an additional physical limitation is applied to optimize the net after the cross-validation method is done. Results show that the neural net inversion method could replace the traditional inversion method in a single formation and can be used a method to determine the initial value of the traditional method. No matter what method is developed, the non-uniqueness and uncertainties of the solution could be inevitable. Thus, it is wise to evaluate the non-uniqueness and uncertainties of the solution in the application of inversion results. Bayes theorem provides a way to solve such problems. This method is illustrately discussed in a single formation and achieve plausible results. In the end, the traditional least squares inversion method is used to process raw logging data, the calculated oil saturation increased 20 percent than that not be proceed compared to core analysis.

Inteligência artificial aplicada na previsão de mercados financeiros

Neste documento, são investigados vários métodos usados na inteligência artificial, com o objetivo de obter previsões precisas da evolução dos mercados financeiros. O uso de ferramentas lineares como os modelos AR, MA, ARMA e GARCH têm muitas limitações, pois torna-se muito difícil adaptá-los às não linearidades dos fenómenos que ocorrem nos mercados. Pelas razões anteriormente referidas, os algoritmos como as redes neuronais dinâmicas (TDNN, NARX e ESN), mostram uma maior capacidade de adaptação a estas não linearidades, pois não fazem qualquer pressuposto sobre as distribuições de probabilidade que caracterizam estes mercados. O facto destas redes neuronais serem dinâmicas, faz com que estas exibam um desempenho superior em relação às redes neuronais estáticas, ou outros algoritmos que não possuem qualquer tipo de memória. Apesar das vantagens reveladas pelas redes neuronais, estas são um sistema do tipo black box, o que torna muito difícil extrair informação dos pesos da rede. Isto significa que estes algoritmos devem ser usados com precaução, pois podem tornar-se instáveis.

Advances in scaling deep learning algorithms

Les algorithmes d'apprentissage profond forment un nouvel ensemble de méthodes puissantes pour l'apprentissage automatique. L'idée est de combiner des couches de facteurs latents en hierarchies. Cela requiert souvent un coût computationel plus elevé et augmente aussi le nombre de paramètres du modèle. Ainsi, l'utilisation de ces méthodes sur des problèmes à plus grande échelle demande de réduire leur coût et aussi d'améliorer leur régularisation et leur optimization. Cette thèse adresse cette question sur ces trois perspectives. Nous étudions tout d'abord le problème de réduire le coût de certains algorithmes profonds. Nous proposons deux méthodes pour entrainer des machines de Boltzmann restreintes et des auto-encodeurs débruitants sur des distributions sparses à haute dimension. Ceci est important pour l'application de ces algorithmes pour le traitement de langues naturelles. Ces deux méthodes (Dauphin et al., 2011; Dauphin and Bengio, 2013) utilisent l'échantillonage par importance pour échantilloner l'objectif de ces modèles. Nous observons que cela réduit significativement le temps d'entrainement. L'accéleration atteint 2 ordres de magnitude sur plusieurs bancs d'essai. Deuxièmement, nous introduisont un puissant régularisateur pour les méthodes profondes. Les résultats expérimentaux démontrent qu'un bon régularisateur est crucial pour obtenir de bonnes performances avec des gros réseaux (Hinton et al., 2012). Dans Rifai et al. (2011), nous proposons un nouveau régularisateur qui combine l'apprentissage non-supervisé et la propagation de tangente (Simard et al., 1992). Cette méthode exploite des principes géometriques et permit au moment de la publication d'atteindre des résultats à l'état de l'art. Finalement, nous considérons le problème d'optimiser des surfaces non-convexes à haute dimensionalité comme celle des réseaux de neurones. Tradionellement, l'abondance de minimum locaux était considéré comme la principale difficulté dans ces problèmes. Dans Dauphin et al. (2014a) nous argumentons à partir de résultats en statistique physique, de la théorie des matrices aléatoires, de la théorie des réseaux de neurones et à partir de résultats expérimentaux qu'une difficulté plus profonde provient de la prolifération de points-selle. Dans ce papier nous proposons aussi une nouvelle méthode pour l'optimisation non-convexe.

FPGA-based object detection using classification circuits

Dans l'apprentissage machine, la classification est le processus d’assigner une nouvelle observation à une certaine catégorie. Les classifieurs qui mettent en œuvre des algorithmes de classification ont été largement étudié au cours des dernières décennies. Les classifieurs traditionnels sont basés sur des algorithmes tels que le SVM et les réseaux de neurones, et sont généralement exécutés par des logiciels sur CPUs qui fait que le système souffre d’un manque de performance et d’une forte consommation d'énergie. Bien que les GPUs puissent être utilisés pour accélérer le calcul de certains classifieurs, leur grande consommation de puissance empêche la technologie d'être mise en œuvre sur des appareils portables tels que les systèmes embarqués. Pour rendre le système de classification plus léger, les classifieurs devraient être capable de fonctionner sur un système matériel plus compact au lieu d'un groupe de CPUs ou GPUs, et les classifieurs eux-mêmes devraient être optimisés pour ce matériel. Dans ce mémoire, nous explorons la mise en œuvre d'un classifieur novateur sur une plate-forme matérielle à base de FPGA. Le classifieur, conçu par Alain Tapp (Université de Montréal), est basé sur une grande quantité de tables de recherche qui forment des circuits arborescents qui effectuent les tâches de classification. Le FPGA semble être un élément fait sur mesure pour mettre en œuvre ce classifieur avec ses riches ressources de tables de recherche et l'architecture à parallélisme élevé. Notre travail montre que les FPGAs peuvent implémenter plusieurs classifieurs et faire les classification sur des images haute définition à une vitesse très élevée.

Los conceptos de salud financiera, riesgo y epidemiología en los estados financieros de las compañías del sector de extracción de petróleo crudo y gas natural en Colombia

El presente proyecto tiene como objeto identificar cuáles son los conceptos de salud, enfermedad, epidemiología y riesgo aplicables a las empresas del sector de extracción de petróleo y gas natural en Colombia. Dado, el bajo nivel de predicción de los análisis financieros tradicionales y su insuficiencia, en términos de inversión y toma de decisiones a largo plazo, además de no considerar variables como el riesgo y las expectativas de futuro, surge la necesidad de abordar diferentes perspectivas y modelos integradores. Esta apreciación es pertinente dentro del sector de extracción de petróleo y gas natural, debido a la creciente inversión extranjera que ha reportado, US$2.862 millones en el 2010, cifra mayor a diez veces su valor en el año 2003. Así pues, se podrían desarrollar modelos multi-dimensional, con base en los conceptos de salud financiera, epidemiológicos y estadísticos. El termino de salud y su adopción en el sector empresarial, resulta útil y mantiene una coherencia conceptual, evidenciando una presencia de diferentes subsistemas o factores interactuantes e interconectados. Es necesario mencionar también, que un modelo multidimensional (multi-stage) debe tener en cuenta el riesgo y el análisis epidemiológico ha demostrado ser útil al momento de determinarlo e integrarlo en el sistema junto a otros conceptos, como la razón de riesgo y riesgo relativo. Esto se analizará mediante un estudio teórico-conceptual, que complementa un estudio previo, para contribuir al proyecto de finanzas corporativas de la línea de investigación en Gerencia.

Synapsing variable length crossover: Biologically inspired crossover for variable length genomes

Biological Crossover occurs during the early stages of meiosis. During this process the chromosomes undergoing crossover are synapsed together at a number of homogenous sequence sections, it is within such synapsed sections that crossover occurs. The SVLC (Synapsing Variable Length Crossover) Algorithm recurrently synapses homogenous genetic sequences together in order of length. The genomes are considered to be flexible with crossover only being permitted within the synapsed sections. Consequently, common sequences are automatically preserved with only the genetic differences being exchanged, independent of the length of such differences. In addition to providing a rationale for variable length crossover it also provides a genotypic similarity metric for variable length genomes enabling standard niche formation techniques to be utilised. In a simple variable length test problem the SVLC algorithm outperforms current variable length crossover techniques.

Simple adaptive momentum: new algorithm for training multilayer perceptrons

The speed of convergence while training is an important consideration in the use of neural nets. The authors outline a new training algorithm which reduces both the number of iterations and training time required for convergence of multilayer perceptrons, compared to standard back-propagation and conjugate gradient descent algorithms.

A novel approach for distributed application scheduling based on prediction of communication events

The evolution of commodity computing lead to the possibility of efficient usage of interconnected machines to solve computationally-intensive tasks, which were previously solvable only by using expensive supercomputers. This, however, required new methods for process scheduling and distribution, considering the network latency, communication cost, heterogeneous environments and distributed computing constraints. An efficient distribution of processes over such environments requires an adequate scheduling strategy, as the cost of inefficient process allocation is unacceptably high. Therefore, a knowledge and prediction of application behavior is essential to perform effective scheduling. In this paper, we overview the evolution of scheduling approaches, focusing on distributed environments. We also evaluate the current approaches for process behavior extraction and prediction, aiming at selecting an adequate technique for online prediction of application execution. Based on this evaluation, we propose a novel model for application behavior prediction, considering chaotic properties of such behavior and the automatic detection of critical execution points. The proposed model is applied and evaluated for process scheduling in cluster and grid computing environments. The obtained results demonstrate that prediction of the process behavior is essential for efficient scheduling in large-scale and heterogeneous distributed environments, outperforming conventional scheduling policies by a factor of 10, and even more in some cases. Furthermore, the proposed approach proves to be efficient for online predictions due to its low computational cost and good precision. (C) 2009 Elsevier B.V. All rights reserved.

Dynamical systems for discovering protein complexes and functional modules from biological networks

Recent advances in high throughput experiments and annotations via published literature have provided a wealth of interaction maps of several biomolecular networks, including metabolic, protein-protein, and protein-DNA interaction networks. The architecture of these molecular networks reveals important principles of cellular organization and molecular functions. Analyzing such networks, i.e., discovering dense regions in the network, is an important way to identify protein complexes and functional modules. This task has been formulated as the problem of finding heavy subgraphs, the Heaviest k-Subgraph Problem (k-HSP), which itself is NPhard. However, any method based on the k-HSP requires the parameter k and an exact solution of k-HSP may still end up as a “spurious” heavy subgraph, thus reducing its practicability in analyzing large scale biological networks. We proposed a new formulation, called the rank-HSP, and two dynamical systems to approximate its results. In addition, a novel metric, called the Standard deviation and Mean Ratio (SMR), is proposed for use in “spurious” heavy subgraphs to automate the discovery by setting a fixed threshold. Empirical results on both the simulated graphs and biological networks have demonstrated the efficiency and effectiveness of our proposal.