A Split-and-Merge Dictionary Learning Algorithm for Sparse Representation: Application to Image Denoising

In big data image/video analytics, we encounter the problem of learning an over-complete dictionary for sparse representation from a large training dataset, which cannot be processed at once because of storage and computational constraints. To tackle the problem of dictionary learning in such scenarios, we propose an algorithm that exploits the inherent clustered structure of the training data and make use of a divide-and-conquer approach. The fundamental idea behind the algorithm is to partition the training dataset into smaller clusters, and learn local dictionaries for each cluster. Subsequently, the local dictionaries are merged to form a global dictionary. Merging is done by solving another dictionary learning problem on the atoms of the locally trained dictionaries. This algorithm is referred to as the split-and-merge algorithm. We show that the proposed algorithm is efficient in its usage of memory and computational complexity, and performs on par with the standard learning strategy, which operates on the entire data at a time. As an application, we consider the problem of image denoising. We present a comparative analysis of our algorithm with the standard learning techniques that use the entire database at a time, in terms of training and denoising performance. We observe that the split-and-merge algorithm results in a remarkable reduction of training time, without significantly affecting the denoising performance.

Características antropométricas, condición física y velocidad de lanzamiento en balonmano de élite y amateur

166 p.

Applications of stochastic context-free grammars using the Inside-Outside algorithm

This paper describes two applications in speech recognition of the use of stochastic context-free grammars (SCFGs) trained automatically via the Inside-Outside Algorithm. First, SCFGs are used to model VQ encoded speech for isolated word recognition and are compared directly to HMMs used for the same task. It is shown that SCFGs can model this low-level VQ data accurately and that a regular grammar based pre-training algorithm is effective both for reducing training time and obtaining robust solutions. Second, an SCFG is inferred from a transcription of the speech used to train a phoneme-based recognizer in an attempt to model phonotactic constraints. When used as a language model, this SCFG gives improved performance over a comparable regular grammar or bigram. © 1991.

Evaluation of a combined wavelet and a combined principal component analysis classification system for BCG diagnostic problem

Heart disease is one of the main factor causing death in the developed countries. Over several decades, variety of electronic and computer technology have been developed to assist clinical practices for cardiac performance monitoring and heart disease diagnosis. Among these methods, Ballistocardiography (BCG) has an interesting feature that no electrodes are needed to be attached to the body during the measurement. Thus, it is provides a potential application to asses the patients heart condition in the home. In this paper, a comparison is made for two neural networks based BCG signal classification models. One system uses a principal component analysis (PCA) method, and the other a discrete wavelet transform, to reduce the input dimensionality. It is indicated that the combined wavelet transform and neural network has a more reliable performance than the combined PCA and neural network system. Moreover, the wavelet transform requires no prior knowledge of the statistical distribution of data samples and the computation complexity and training time are reduced.

Sharing visual features for multiclass and multiview object detection

We consider the problem of detecting a large number of different classes of objects in cluttered scenes. Traditional approaches require applying a battery of different classifiers to the image, at multiple locations and scales. This can be slow and can require a lot of training data, since each classifier requires the computation of many different image features. In particular, for independently trained detectors, the (run-time) computational complexity, and the (training-time) sample complexity, scales linearly with the number of classes to be detected. It seems unlikely that such an approach will scale up to allow recognition of hundreds or thousands of objects. We present a multi-class boosting procedure (joint boosting) that reduces the computational and sample complexity, by finding common features that can be shared across the classes (and/or views). The detectors for each class are trained jointly, rather than independently. For a given performance level, the total number of features required, and therefore the computational cost, is observed to scale approximately logarithmically with the number of classes. The features selected jointly are closer to edges and generic features typical of many natural structures instead of finding specific object parts. Those generic features generalize better and reduce considerably the computational cost of an algorithm for multi-class object detection.

Improved conjugate gradient implementation for least squares support vector machines

As a promising method for pattern recognition and function estimation, least squares support vector machines (LS-SVM) express the training in terms of solving a linear system instead of a quadratic programming problem as for conventional support vector machines (SVM). In this paper, by using the information provided by the equality constraint, we transform the minimization problem with a single equality constraint in LS-SVM into an unconstrained minimization problem, then propose reduced formulations for LS-SVM. By introducing this transformation, the times of using conjugate gradient (CG) method, which is a greatly time-consuming step in obtaining the numerical solution, are reduced to one instead of two as proposed by Suykens et al. (1999). The comparison on computational speed of our method with the CG method proposed by Suykens et al. and the first order and second order SMO methods on several benchmark data sets shows a reduction of training time by up to 44%. (C) 2011 Elsevier B.V. All rights reserved.

Scalable black-box prediction models for multi-dimensional adaptation on NUMA multi-cores

This paper presents a scalable, statistical ‘black-box’ model for predicting the performance of parallel programs on multi-core non-uniform memory access (NUMA) systems. We derive a model with low overhead, by reducing data collection and model training time. The model can accurately predict the behaviour of parallel applications in response to changes in their concurrency, thread layout on NUMA nodes, and core voltage and frequency. We present a framework that applies the model to achieve significant energy and energy-delay-square (ED2) savings (9% and 25%, respectively) along with performance improvement (10% mean) on an actual 16-core NUMA system running realistic application workloads. Our prediction model proves substantially more accurate than previous efforts.

O Contributo do sistema SAP R/3 no controlo interno na REFER

Dissertação apresentada ao Instituto Superior de Contabilidade para a obtenção do Grau de Mestre em Auditoria Orientador: Mestre Agostinho Sousa Pinto

Safe Handling of Large Animals

The article discusses improving welfare by reducing fear by studying: Animal Sensory Perception, Animal Behavior Patterns, Animal Habituation and Temperament, Effects of Previous Handling, Training Animals, Training Time and Temperament, Genetic Effects on Handling, Handling of escaped Animals, Facilities, Aggression in Grazing Animals, Inherent Danger of Large Animals, Cattle and Car Accidents.

L'effet de l'entrainement avec le ballon d'exercice pendant la grossesse sur l'accouchement.

Résumé Introduction : L’entrainement avec le ballon d’exercice pendant la grossesse a des effets positifs sur le déroulement de l’accouchement. Le but de l’étude était d’identifier si une association existe entre le volume d’entrainement avec le ballon d’exercice durant la grossesse en milieu naturel sur l’issue du déroulement de l’accouchement, soit la durée des phases 1 (dilatation/phase active) et 2 (expulsion), ainsi que l’usage de la péridurale et de la césarienne. Méthode: Cette étude quasi-expérimentale a été menée auprès d'un échantillon de convenance composée de 32 femmes enceintes qui ont pris part entre un et 28 cours d'exercice supervisé durant la grossesse (programme Ballon Forme), avec la possibilité d'effectuer des exercices supplémentaires à la maison avec le ballon d’exercice. Un questionnaire et un journal de bord ont été remplis par les participantes, avec l'aide d'une infirmière de l’hôpital pour les données médicales. Résultats: Un volume d’entrainement élevé, incluant l’entraînement sous forme de cours et à domicile, a été significativement associé à une durée plus courte d’accouchement, que ce soit pour le temps total de l’accouchement (r = -0,408, p = 0,031) ou pour les deux phases spécifiques de l’accouchement [la phase 1: r = -0,372 ; p = 0,043 et la phase 2: r = -0,415, p = 0,028]. Un volume d’entrainement élevé a aussi été lié à une réduction de la deuxième phase chez les femmes qui donnaient naissance pour la première fois. La durée totale des exercices exécutés lors des cours a été significativement associée à cette réduction comparativement aux exercices à la maison. Aucun effet indésirable n'a été observé avec le volume d’entrainement élevé sur les paramètres de l’accouchement et la santé du bébé (Apgar et poids du bébé à la naissance). En fait, un nombre très faible de césarienne (6%), de recours à la péridurale (47%) et d'extraction instrumentale (forceps ou ventouse: 13%) a été observé dans notre échantillon de femmes. Conclusion: La pratique du ballon d’exercice est une avenue prometteuse pour les femmes enceintes et un niveau élevé d’entraînement est associé à un accouchement plus rapide et sans complications pour le bébé. L’implantation de ce programme pourrait éventuellement contribuer à réduire considérablement les coûts de santé au Québec en favorisant l’accouchement naturel. Des interventions et des recherches ultérieures devront considérer et examiner cette modalité spécifique d’entrainement pour les femmes enceintes.

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.

Face Recognition using Probabilistic Neural Networks

In this paper we address the problem of face detection and recognition of grey scale frontal view images. We propose a face recognition system based on probabilistic neural networks (PNN) architecture. The system is implemented using voronoi/ delaunay tessellations and template matching. Images are segmented successfully into homogeneous regions by virtue of voronoi diagram properties. Face verification is achieved using matching scores computed by correlating edge gradients of reference images. The advantage of classification using PNN models is its short training time. The correlation based template matching guarantees good classification results

Face Recognition using Probabilistic Neural Networks

n this paper we address the problem of face detection and recognition of grey scale frontal view images. We propose a face recognition system based on probabilistic neural networks (PNN) architecture. The system is implemented using voronoi/ delaunay tessellations and template matching. Images are segmented successfully into homogeneous regions by virtue of voronoi diagram properties. Face verification is achieved using matching scores computed by correlating edge gradients of reference images. The advantage of classification using PNN models is its short training time. The correlation based template matching guarantees good classification results.

Increasing detection rate of user-to-root attacks using genetic algorithms

An extensive set of machine learning and pattern classification techniques trained and tested on KDD dataset failed in detecting most of the user-to-root attacks. This paper aims to provide an approach for mitigating negative aspects of the mentioned dataset, which led to low detection rates. Genetic algorithm is employed to implement rules for detecting various types of attacks. Rules are formed of the features of the dataset identified as the most important ones for each attack type. In this way we introduce high level of generality and thus achieve high detection rates, but also gain high reduction of the system training time. Thenceforth we re-check the decision of the user-to- root rules with the rules that detect other types of attacks. In this way we decrease the false-positive rate. The model was verified on KDD 99, demonstrating higher detection rates than those reported by the state- of-the-art while maintaining low false-positive rate.

An Information Theory framework for two-stage binary image operator design

The design of translation invariant and locally defined binary image operators over large windows is made difficult by decreased statistical precision and increased training time. We present a complete framework for the application of stacked design, a recently proposed technique to create two-stage operators that circumvents that difficulty. We propose a novel algorithm, based on Information Theory, to find groups of pixels that should be used together to predict the Output Value. We employ this algorithm to automate the process of creating a set of first-level operators that are later combined in a global operator. We also propose a principled way to guide this combination, by using feature selection and model comparison. Experimental results Show that the proposed framework leads to better results than single stage design. (C) 2009 Elsevier B.V. All rights reserved.