On Automatic Diagnosis of Alzheimer's Disease based on Spontaneous Speech Analysis and Emotional Temperature

Alzheimer's disease is the most prevalent form of progressive degenerative dementia; it has a high socio-economic impact in Western countries. Therefore it is one of the most active research areas today. Alzheimer's is sometimes diagnosed by excluding other dementias, and definitive confirmation is only obtained through a post-mortem study of the brain tissue of the patient. The work presented here is part of a larger study that aims to identify novel technologies and biomarkers for early Alzheimer's disease detection, and it focuses on evaluating the suitability of a new approach for early diagnosis of Alzheimer’s disease by non-invasive methods. The purpose is to examine, in a pilot study, the potential of applying Machine Learning algorithms to speech features obtained from suspected Alzheimer sufferers in order help diagnose this disease and determine its degree of severity. Two human capabilities relevant in communication have been analyzed for feature selection: Spontaneous Speech and Emotional Response. The experimental results obtained were very satisfactory and promising for the early diagnosis and classification of Alzheimer’s disease patients.

Modelling ecological niches with support vector machines

1. The ecological niche is a fundamental biological concept. Modelling species' niches is central to numerous ecological applications, including predicting species invasions, identifying reservoirs for disease, nature reserve design and forecasting the effects of anthropogenic and natural climate change on species' ranges. 2. A computational analogue of Hutchinson's ecological niche concept (the multidimensional hyperspace of species' environmental requirements) is the support of the distribution of environments in which the species persist. Recently developed machine-learning algorithms can estimate the support of such high-dimensional distributions. We show how support vector machines can be used to map ecological niches using only observations of species presence to train distribution models for 106 species of woody plants and trees in a montane environment using up to nine environmental covariates. 3. We compared the accuracy of three methods that differ in their approaches to reducing model complexity. We tested models with independent observations of both species presence and species absence. We found that the simplest procedure, which uses all available variables and no pre-processing to reduce correlation, was best overall. Ecological niche models based on support vector machines are theoretically superior to models that rely on simulating pseudo-absence data and are comparable in empirical tests. 4. Synthesis and applications. Accurate species distribution models are crucial for effective environmental planning, management and conservation, and for unravelling the role of the environment in human health and welfare. Models based on distribution estimation rather than classification overcome theoretical and practical obstacles that pervade species distribution modelling. In particular, ecological niche models based on machine-learning algorithms for estimating the support of a statistical distribution provide a promising new approach to identifying species' potential distributions and to project changes in these distributions as a result of climate change, land use and landscape alteration.

Robust Nonlinear Mapping of Soil Contamination Using Support Regression

Spatial data analysis mapping and visualization is of great importance in various fields: environment, pollution, natural hazards and risks, epidemiology, spatial econometrics, etc. A basic task of spatial mapping is to make predictions based on some empirical data (measurements). A number of state-of-the-art methods can be used for the task: deterministic interpolations, methods of geostatistics: the family of kriging estimators (Deutsch and Journel, 1997), machine learning algorithms such as artificial neural networks (ANN) of different architectures, hybrid ANN-geostatistics models (Kanevski and Maignan, 2004; Kanevski et al., 1996), etc. All the methods mentioned above can be used for solving the problem of spatial data mapping. Environmental empirical data are always contaminated/corrupted by noise, and often with noise of unknown nature. That's one of the reasons why deterministic models can be inconsistent, since they treat the measurements as values of some unknown function that should be interpolated. Kriging estimators treat the measurements as the realization of some spatial randomn process. To obtain the estimation with kriging one has to model the spatial structure of the data: spatial correlation function or (semi-)variogram. This task can be complicated if there is not sufficient number of measurements and variogram is sensitive to outliers and extremes. ANN is a powerful tool, but it also suffers from the number of reasons. of a special type ? multiplayer perceptrons ? are often used as a detrending tool in hybrid (ANN+geostatistics) models (Kanevski and Maignank, 2004). Therefore, development and adaptation of the method that would be nonlinear and robust to noise in measurements, would deal with the small empirical datasets and which has solid mathematical background is of great importance. The present paper deals with such model, based on Statistical Learning Theory (SLT) - Support Vector Regression. SLT is a general mathematical framework devoted to the problem of estimation of the dependencies from empirical data (Hastie et al, 2004; Vapnik, 1998). SLT models for classification - Support Vector Machines - have shown good results on different machine learning tasks. The results of SVM classification of spatial data are also promising (Kanevski et al, 2002). The properties of SVM for regression - Support Vector Regression (SVR) are less studied. First results of the application of SVR for spatial mapping of physical quantities were obtained by the authorsin for mapping of medium porosity (Kanevski et al, 1999), and for mapping of radioactively contaminated territories (Kanevski and Canu, 2000). The present paper is devoted to further understanding of the properties of SVR model for spatial data analysis and mapping. Detailed description of the SVR theory can be found in (Cristianini and Shawe-Taylor, 2000; Smola, 1996) and basic equations for the nonlinear modeling are given in section 2. Section 3 discusses the application of SVR for spatial data mapping on the real case study - soil pollution by Cs137 radionuclide. Section 4 discusses the properties of the modelapplied to noised data or data with outliers.

Interest rates mapping

The present study deals with the analysis and mapping of Swiss franc interest rates. Interest rates depend on time and maturity, defining term structure of the interest rate curves (IRC). In the present study IRC are considered in a two-dimensional feature space - time and maturity. Exploratory data analysis includes a variety of tools widely used in econophysics and geostatistics. Geostatistical models and machine learning algorithms (multilayer perceptron and Support Vector Machines) were applied to produce interest rate maps. IR maps can be used for the visualisation and pattern perception purposes, to develop and to explore economical hypotheses, to produce dynamic asset-liability simulations and for financial risk assessments. The feasibility of an application of interest rates mapping approach for the IRC forecasting is considered as well. (C) 2008 Elsevier B.V. All rights reserved.

Scalable Feature Selection Applications for Genome-Wide Association Studies of Complex Diseases

Personalized medicine will revolutionize our capabilities to combat disease. Working toward this goal, a fundamental task is the deciphering of geneticvariants that are predictive of complex diseases. Modern studies, in the formof genome-wide association studies (GWAS) have afforded researchers with the opportunity to reveal new genotype-phenotype relationships through the extensive scanning of genetic variants. These studies typically contain over half a million genetic features for thousands of individuals. Examining this with methods other than univariate statistics is a challenging task requiring advanced algorithms that are scalable to the genome-wide level. In the future, next-generation sequencing studies (NGS) will contain an even larger number of common and rare variants. Machine learning-based feature selection algorithms have been shown to have the ability to effectively create predictive models for various genotype-phenotype relationships. This work explores the problem of selecting genetic variant subsets that are the most predictive of complex disease phenotypes through various feature selection methodologies, including filter, wrapper and embedded algorithms. The examined machine learning algorithms were demonstrated to not only be effective at predicting the disease phenotypes, but also doing so efficiently through the use of computational shortcuts. While much of the work was able to be run on high-end desktops, some work was further extended so that it could be implemented on parallel computers helping to assure that they will also scale to the NGS data sets. Further, these studies analyzed the relationships between various feature selection methods and demonstrated the need for careful testing when selecting an algorithm. It was shown that there is no universally optimal algorithm for variant selection in GWAS, but rather methodologies need to be selected based on the desired outcome, such as the number of features to be included in the prediction model. It was also demonstrated that without proper model validation, for example using nested cross-validation, the models can result in overly-optimistic prediction accuracies and decreased generalization ability. It is through the implementation and application of machine learning methods that one can extract predictive genotype–phenotype relationships and biological insights from genetic data sets.

Analyse par apprentissage automatique des réponses fMRI du cortex auditif à des modulations spectro-temporelles

L'application de classifieurs linéaires à l'analyse des données d'imagerie cérébrale (fMRI) a mené à plusieurs percées intéressantes au cours des dernières années. Ces classifieurs combinent linéairement les réponses des voxels pour détecter et catégoriser différents états du cerveau. Ils sont plus agnostics que les méthodes d'analyses conventionnelles qui traitent systématiquement les patterns faibles et distribués comme du bruit. Dans le présent projet, nous utilisons ces classifieurs pour valider une hypothèse portant sur l'encodage des sons dans le cerveau humain. Plus précisément, nous cherchons à localiser des neurones, dans le cortex auditif primaire, qui détecteraient les modulations spectrales et temporelles présentes dans les sons. Nous utilisons les enregistrements fMRI de sujets soumis à 49 modulations spectro-temporelles différentes. L'analyse fMRI au moyen de classifieurs linéaires n'est pas standard, jusqu'à maintenant, dans ce domaine. De plus, à long terme, nous avons aussi pour objectif le développement de nouveaux algorithmes d'apprentissage automatique spécialisés pour les données fMRI. Pour ces raisons, une bonne partie des expériences vise surtout à étudier le comportement des classifieurs. Nous nous intéressons principalement à 3 classifieurs linéaires standards, soient l'algorithme machine à vecteurs de support (linéaire), l'algorithme régression logistique (régularisée) et le modèle bayésien gaussien naïf (variances partagées).

Espaces de timbre générés par des réseaux profonds convolutionnels

Il est avant-tout question, dans ce mémoire, de la modélisation du timbre grâce à des algorithmes d'apprentissage machine. Plus précisément, nous avons essayé de construire un espace de timbre en extrayant des caractéristiques du son à l'aide de machines de Boltzmann convolutionnelles profondes. Nous présentons d'abord un survol de l'apprentissage machine, avec emphase sur les machines de Boltzmann convolutionelles ainsi que les modèles dont elles sont dérivées. Nous présentons aussi un aperçu de la littérature concernant les espaces de timbre, et mettons en évidence quelque-unes de leurs limitations, dont le nombre limité de sons utilisés pour les construire. Pour pallier à ce problème, nous avons mis en place un outil nous permettant de générer des sons à volonté. Le système utilise à sa base des plug-ins qu'on peut combiner et dont on peut changer les paramètres pour créer une gamme virtuellement infinie de sons. Nous l'utilisons pour créer une gigantesque base de donnée de timbres générés aléatoirement constituée de vrais instruments et d'instruments synthétiques. Nous entrainons ensuite les machines de Boltzmann convolutionnelles profondes de façon non-supervisée sur ces timbres, et utilisons l'espace des caractéristiques produites comme espace de timbre. L'espace de timbre ainsi obtenu est meilleur qu'un espace semblable construit à l'aide de MFCC. Il est meilleur dans le sens où la distance entre deux timbres dans cet espace est plus semblable à celle perçue par un humain. Cependant, nous sommes encore loin d'atteindre les mêmes capacités qu'un humain. Nous proposons d'ailleurs quelques pistes d'amélioration pour s'en approcher.

Évaluation et amélioration du rendement de la formation en entreprise : vers une démarche basée sur la gestion des processus d’affaires.

La formation est une stratégie clé pour le développement des compétences. Les entreprises continuent à investir dans la formation et le développement, mais elles possèdent rarement des données pour évaluer les résultats de cet investissement. La plupart des entreprises utilisent le modèle Kirkpatrick/Phillips pour évaluer la formation en entreprise. Cependant, il ressort de la littérature que les entreprises ont des difficultés à utiliser ce modèle. Les principales barrières sont la difficulté d’isoler l’apprentissage comme un facteur qui a une incidence sur les résultats, l’absence d’un système d’évaluation utile avec le système de gestion de l’apprentissage (Learning Management System - LMS) et le manque de données standardisées pour pouvoir comparer différentes fonctions d’apprentissage. Dans cette thèse, nous proposons un modèle (Analyse, Modélisation, Monitoring et Optimisation - AM2O) de gestion de projets de formation en entreprise, basée sur la gestion des processus d’affaires (Business Process Management - BPM). Un tel scénario suppose que les activités de formation en entreprise doivent être considérées comme des processus d’affaires. Notre modèle est inspiré de cette méthode (BPM), à travers la définition et le suivi des indicateurs de performance pour gérer les projets de formation dans les organisations. Elle est basée sur l’analyse et la modélisation des besoins de formation pour assurer l’alignement entre les activités de formation et les objectifs d’affaires de l’entreprise. Elle permet le suivi des projets de formation ainsi que le calcul des avantages tangibles et intangibles de la formation (sans coût supplémentaire). En outre, elle permet la production d’une classification des projets de formation en fonction de critères relatifs à l’entreprise. Ainsi, avec assez de données, notre approche peut être utilisée pour optimiser le rendement de la formation par une série de simulations utilisant des algorithmes d’apprentissage machine : régression logistique, réseau de neurones, co-apprentissage. Enfin, nous avons conçu un système informatique, Enterprise TRaining programs Evaluation and Optimization System - ETREOSys, pour la gestion des programmes de formation en entreprise et l’aide à la décision. ETREOSys est une plateforme Web utilisant des services en nuage (cloud services) et les bases de données NoSQL. A travers AM2O et ETREOSys nous résolvons les principaux problèmes liés à la gestion et l’évaluation de la formation en entreprise à savoir : la difficulté d’isoler les effets de la formation dans les résultats de l’entreprise et le manque de systèmes informatiques.

Apprentissage d'espaces sémantiques

Dans cette dissertation, nous présentons plusieurs techniques d’apprentissage d’espaces sémantiques pour plusieurs domaines, par exemple des mots et des images, mais aussi à l’intersection de différents domaines. Un espace de représentation est appelé sémantique si des entités jugées similaires par un être humain, ont leur similarité préservée dans cet espace. La première publication présente un enchaînement de méthodes d’apprentissage incluant plusieurs techniques d’apprentissage non supervisé qui nous a permis de remporter la compétition “Unsupervised and Transfer Learning Challenge” en 2011. Le deuxième article présente une manière d’extraire de l’information à partir d’un contexte structuré (177 détecteurs d’objets à différentes positions et échelles). On montrera que l’utilisation de la structure des données combinée à un apprentissage non supervisé permet de réduire la dimensionnalité de 97% tout en améliorant les performances de reconnaissance de scènes de +5% à +11% selon l’ensemble de données. Dans le troisième travail, on s’intéresse à la structure apprise par les réseaux de neurones profonds utilisés dans les deux précédentes publications. Plusieurs hypothèses sont présentées et testées expérimentalement montrant que l’espace appris a de meilleures propriétés de mixage (facilitant l’exploration de différentes classes durant le processus d’échantillonnage). Pour la quatrième publication, on s’intéresse à résoudre un problème d’analyse syntaxique et sémantique avec des réseaux de neurones récurrents appris sur des fenêtres de contexte de mots. Dans notre cinquième travail, nous proposons une façon d’effectuer de la recherche d’image ”augmentée” en apprenant un espace sémantique joint où une recherche d’image contenant un objet retournerait aussi des images des parties de l’objet, par exemple une recherche retournant des images de ”voiture” retournerait aussi des images de ”pare-brises”, ”coffres”, ”roues” en plus des images initiales.

A review on the combination of binary classifiers in multiclass problems

Several real problems involve the classification of data into categories or classes. Given a data set containing data whose classes are known, Machine Learning algorithms can be employed for the induction of a classifier able to predict the class of new data from the same domain, performing the desired discrimination. Some learning techniques are originally conceived for the solution of problems with only two classes, also named binary classification problems. However, many problems require the discrimination of examples into more than two categories or classes. This paper presents a survey on the main strategies for the generalization of binary classifiers to problems with more than two classes, known as multiclass classification problems. The focus is on strategies that decompose the original multiclass problem into multiple binary subtasks, whose outputs are combined to obtain the final prediction.

A hybrid case adaptation approach for case-based reasoning

Case-Based Reasoning is a methodology for problem solving based on past experiences. This methodology tries to solve a new problem by retrieving and adapting previously known solutions of similar problems. However, retrieved solutions, in general, require adaptations in order to be applied to new contexts. One of the major challenges in Case-Based Reasoning is the development of an efficient methodology for case adaptation. The most widely used form of adaptation employs hand coded adaptation rules, which demands a significant knowledge acquisition and engineering effort. An alternative to overcome the difficulties associated with the acquisition of knowledge for case adaptation has been the use of hybrid approaches and automatic learning algorithms for the acquisition of the knowledge used for the adaptation. We investigate the use of hybrid approaches for case adaptation employing Machine Learning algorithms. The approaches investigated how to automatically learn adaptation knowledge from a case base and apply it to adapt retrieved solutions. In order to verify the potential of the proposed approaches, they are experimentally compared with individual Machine Learning techniques. The results obtained indicate the potential of these approaches as an efficient approach for acquiring case adaptation knowledge. They show that the combination of Instance-Based Learning and Inductive Learning paradigms and the use of a data set of adaptation patterns yield adaptations of the retrieved solutions with high predictive accuracy.

A path- and label-cost propagation approach to speedup the training of the optimum-path forest classifier

In general, pattern recognition techniques require a high computational burden for learning the discriminating functions that are responsible to separate samples from distinct classes. As such, there are several studies that make effort to employ machine learning algorithms in the context of big data classification problems. The research on this area ranges from Graphics Processing Units-based implementations to mathematical optimizations, being the main drawback of the former approaches to be dependent on the graphic video card. Here, we propose an architecture-independent optimization approach for the optimum-path forest (OPF) classifier, that is designed using a theoretical formulation that relates the minimum spanning tree with the minimum spanning forest generated by the OPF over the training dataset. The experiments have shown that the approach proposed can be faster than the traditional one in five public datasets, being also as accurate as the original OPF. (C) 2014 Elsevier B. V. All rights reserved.

Explorando abordagens de múltiplos rótulos por floresta de caminhos ótimos

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Segmentação interativa de imagens utilizando competição e cooperação entre partículas

Pós-graduação em Ciência da Computação - IBILCE

Classificazione di oggetti in immagini attraverso il modello Bag of Visual Words

Generic object recognition is an important function of the human visual system and everybody finds it highly useful in their everyday life. For an artificial vision system it is a really hard, complex and challenging task because instances of the same object category can generate very different images, depending of different variables such as illumination conditions, the pose of an object, the viewpoint of the camera, partial occlusions, and unrelated background clutter. The purpose of this thesis is to develop a system that is able to classify objects in 2D images based on the context, and identify to which category the object belongs to. Given an image, the system can classify it and decide the correct categorie of the object. Furthermore the objective of this thesis is also to test the performance and the precision of different supervised Machine Learning algorithms in this specific task of object image categorization. Through different experiments the implemented application reveals good categorization performances despite the difficulty of the problem. However this project is open to future improvement; it is possible to implement new algorithms that has not been invented yet or using other techniques to extract features to make the system more reliable. This application can be installed inside an embedded system and after trained (performed outside the system), so it can become able to classify objects in a real-time. The information given from a 3D stereocamera, developed inside the department of Computer Engineering of the University of Bologna, can be used to improve the accuracy of the classification task. The idea is to segment a single object in a scene using the depth given from a stereocamera and in this way make the classification more accurate.