Contributions to Bayesian network learning with applications to neuroscience

Neuronal morphology is a key feature in the study of brain circuits, as it is highly related to information processing and functional identification. Neuronal morphology affects the process of integration of inputs from other neurons and determines the neurons which receive the output of the neurons. Different parts of the neurons can operate semi-independently according to the spatial location of the synaptic connections. As a result, there is considerable interest in the analysis of the microanatomy of nervous cells since it constitutes an excellent tool for better understanding cortical function. However, the morphologies, molecular features and electrophysiological properties of neuronal cells are extremely variable. Except for some special cases, this variability makes it hard to find a set of features that unambiguously define a neuronal type. In addition, there are distinct types of neurons in particular regions of the brain. This morphological variability makes the analysis and modeling of neuronal morphology a challenge. Uncertainty is a key feature in many complex real-world problems. Probability theory provides a framework for modeling and reasoning with uncertainty. Probabilistic graphical models combine statistical theory and graph theory to provide a tool for managing domains with uncertainty. In particular, we focus on Bayesian networks, the most commonly used probabilistic graphical model. In this dissertation, we design new methods for learning Bayesian networks and apply them to the problem of modeling and analyzing morphological data from neurons. The morphology of a neuron can be quantified using a number of measurements, e.g., the length of the dendrites and the axon, the number of bifurcations, the direction of the dendrites and the axon, etc. These measurements can be modeled as discrete or continuous data. The continuous data can be linear (e.g., the length or the width of a dendrite) or directional (e.g., the direction of the axon). These data may follow complex probability distributions and may not fit any known parametric distribution. Modeling this kind of problems using hybrid Bayesian networks with discrete, linear and directional variables poses a number of challenges regarding learning from data, inference, etc. In this dissertation, we propose a method for modeling and simulating basal dendritic trees from pyramidal neurons using Bayesian networks to capture the interactions between the variables in the problem domain. A complete set of variables is measured from the dendrites, and a learning algorithm is applied to find the structure and estimate the parameters of the probability distributions included in the Bayesian networks. Then, a simulation algorithm is used to build the virtual dendrites by sampling values from the Bayesian networks, and a thorough evaluation is performed to show the model’s ability to generate realistic dendrites. In this first approach, the variables are discretized so that discrete Bayesian networks can be learned and simulated. Then, we address the problem of learning hybrid Bayesian networks with different kinds of variables. Mixtures of polynomials have been proposed as a way of representing probability densities in hybrid Bayesian networks. We present a method for learning mixtures of polynomials approximations of one-dimensional, multidimensional and conditional probability densities from data. The method is based on basis spline interpolation, where a density is approximated as a linear combination of basis splines. The proposed algorithms are evaluated using artificial datasets. We also use the proposed methods as a non-parametric density estimation technique in Bayesian network classifiers. Next, we address the problem of including directional data in Bayesian networks. These data have some special properties that rule out the use of classical statistics. Therefore, different distributions and statistics, such as the univariate von Mises and the multivariate von Mises–Fisher distributions, should be used to deal with this kind of information. In particular, we extend the naive Bayes classifier to the case where the conditional probability distributions of the predictive variables given the class follow either of these distributions. We consider the simple scenario, where only directional predictive variables are used, and the hybrid case, where discrete, Gaussian and directional distributions are mixed. The classifier decision functions and their decision surfaces are studied at length. Artificial examples are used to illustrate the behavior of the classifiers. The proposed classifiers are empirically evaluated over real datasets. We also study the problem of interneuron classification. An extensive group of experts is asked to classify a set of neurons according to their most prominent anatomical features. A web application is developed to retrieve the experts’ classifications. We compute agreement measures to analyze the consensus between the experts when classifying the neurons. Using Bayesian networks and clustering algorithms on the resulting data, we investigate the suitability of the anatomical terms and neuron types commonly used in the literature. Additionally, we apply supervised learning approaches to automatically classify interneurons using the values of their morphological measurements. Then, a methodology for building a model which captures the opinions of all the experts is presented. First, one Bayesian network is learned for each expert, and we propose an algorithm for clustering Bayesian networks corresponding to experts with similar behaviors. Then, a Bayesian network which represents the opinions of each group of experts is induced. Finally, a consensus Bayesian multinet which models the opinions of the whole group of experts is built. A thorough analysis of the consensus model identifies different behaviors between the experts when classifying the interneurons in the experiment. A set of characterizing morphological traits for the neuronal types can be defined by performing inference in the Bayesian multinet. These findings are used to validate the model and to gain some insights into neuron morphology. Finally, we study a classification problem where the true class label of the training instances is not known. Instead, a set of class labels is available for each instance. This is inspired by the neuron classification problem, where a group of experts is asked to individually provide a class label for each instance. We propose a novel approach for learning Bayesian networks using count vectors which represent the number of experts who selected each class label for each instance. These Bayesian networks are evaluated using artificial datasets from supervised learning problems. Resumen La morfología neuronal es una característica clave en el estudio de los circuitos cerebrales, ya que está altamente relacionada con el procesado de información y con los roles funcionales. La morfología neuronal afecta al proceso de integración de las señales de entrada y determina las neuronas que reciben las salidas de otras neuronas. Las diferentes partes de la neurona pueden operar de forma semi-independiente de acuerdo a la localización espacial de las conexiones sinápticas. Por tanto, existe un interés considerable en el análisis de la microanatomía de las células nerviosas, ya que constituye una excelente herramienta para comprender mejor el funcionamiento de la corteza cerebral. Sin embargo, las propiedades morfológicas, moleculares y electrofisiológicas de las células neuronales son extremadamente variables. Excepto en algunos casos especiales, esta variabilidad morfológica dificulta la definición de un conjunto de características que distingan claramente un tipo neuronal. Además, existen diferentes tipos de neuronas en regiones particulares del cerebro. La variabilidad neuronal hace que el análisis y el modelado de la morfología neuronal sean un importante reto científico. La incertidumbre es una propiedad clave en muchos problemas reales. La teoría de la probabilidad proporciona un marco para modelar y razonar bajo incertidumbre. Los modelos gráficos probabilísticos combinan la teoría estadística y la teoría de grafos con el objetivo de proporcionar una herramienta con la que trabajar bajo incertidumbre. En particular, nos centraremos en las redes bayesianas, el modelo más utilizado dentro de los modelos gráficos probabilísticos. En esta tesis hemos diseñado nuevos métodos para aprender redes bayesianas, inspirados por y aplicados al problema del modelado y análisis de datos morfológicos de neuronas. La morfología de una neurona puede ser cuantificada usando una serie de medidas, por ejemplo, la longitud de las dendritas y el axón, el número de bifurcaciones, la dirección de las dendritas y el axón, etc. Estas medidas pueden ser modeladas como datos continuos o discretos. A su vez, los datos continuos pueden ser lineales (por ejemplo, la longitud o la anchura de una dendrita) o direccionales (por ejemplo, la dirección del axón). Estos datos pueden llegar a seguir distribuciones de probabilidad muy complejas y pueden no ajustarse a ninguna distribución paramétrica conocida. El modelado de este tipo de problemas con redes bayesianas híbridas incluyendo variables discretas, lineales y direccionales presenta una serie de retos en relación al aprendizaje a partir de datos, la inferencia, etc. En esta tesis se propone un método para modelar y simular árboles dendríticos basales de neuronas piramidales usando redes bayesianas para capturar las interacciones entre las variables del problema. Para ello, se mide un amplio conjunto de variables de las dendritas y se aplica un algoritmo de aprendizaje con el que se aprende la estructura y se estiman los parámetros de las distribuciones de probabilidad que constituyen las redes bayesianas. Después, se usa un algoritmo de simulación para construir dendritas virtuales mediante el muestreo de valores de las redes bayesianas. Finalmente, se lleva a cabo una profunda evaluaci ón para verificar la capacidad del modelo a la hora de generar dendritas realistas. En esta primera aproximación, las variables fueron discretizadas para poder aprender y muestrear las redes bayesianas. A continuación, se aborda el problema del aprendizaje de redes bayesianas con diferentes tipos de variables. Las mixturas de polinomios constituyen un método para representar densidades de probabilidad en redes bayesianas híbridas. Presentamos un método para aprender aproximaciones de densidades unidimensionales, multidimensionales y condicionales a partir de datos utilizando mixturas de polinomios. El método se basa en interpolación con splines, que aproxima una densidad como una combinación lineal de splines. Los algoritmos propuestos se evalúan utilizando bases de datos artificiales. Además, las mixturas de polinomios son utilizadas como un método no paramétrico de estimación de densidades para clasificadores basados en redes bayesianas. Después, se estudia el problema de incluir información direccional en redes bayesianas. Este tipo de datos presenta una serie de características especiales que impiden el uso de las técnicas estadísticas clásicas. Por ello, para manejar este tipo de información se deben usar estadísticos y distribuciones de probabilidad específicos, como la distribución univariante von Mises y la distribución multivariante von Mises–Fisher. En concreto, en esta tesis extendemos el clasificador naive Bayes al caso en el que las distribuciones de probabilidad condicionada de las variables predictoras dada la clase siguen alguna de estas distribuciones. Se estudia el caso base, en el que sólo se utilizan variables direccionales, y el caso híbrido, en el que variables discretas, lineales y direccionales aparecen mezcladas. También se estudian los clasificadores desde un punto de vista teórico, derivando sus funciones de decisión y las superficies de decisión asociadas. El comportamiento de los clasificadores se ilustra utilizando bases de datos artificiales. Además, los clasificadores son evaluados empíricamente utilizando bases de datos reales. También se estudia el problema de la clasificación de interneuronas. Desarrollamos una aplicación web que permite a un grupo de expertos clasificar un conjunto de neuronas de acuerdo a sus características morfológicas más destacadas. Se utilizan medidas de concordancia para analizar el consenso entre los expertos a la hora de clasificar las neuronas. Se investiga la idoneidad de los términos anatómicos y de los tipos neuronales utilizados frecuentemente en la literatura a través del análisis de redes bayesianas y la aplicación de algoritmos de clustering. Además, se aplican técnicas de aprendizaje supervisado con el objetivo de clasificar de forma automática las interneuronas a partir de sus valores morfológicos. A continuación, se presenta una metodología para construir un modelo que captura las opiniones de todos los expertos. Primero, se genera una red bayesiana para cada experto y se propone un algoritmo para agrupar las redes bayesianas que se corresponden con expertos con comportamientos similares. Después, se induce una red bayesiana que modela la opinión de cada grupo de expertos. Por último, se construye una multired bayesiana que modela las opiniones del conjunto completo de expertos. El análisis del modelo consensuado permite identificar diferentes comportamientos entre los expertos a la hora de clasificar las neuronas. Además, permite extraer un conjunto de características morfológicas relevantes para cada uno de los tipos neuronales mediante inferencia con la multired bayesiana. Estos descubrimientos se utilizan para validar el modelo y constituyen información relevante acerca de la morfología neuronal. Por último, se estudia un problema de clasificación en el que la etiqueta de clase de los datos de entrenamiento es incierta. En cambio, disponemos de un conjunto de etiquetas para cada instancia. Este problema está inspirado en el problema de la clasificación de neuronas, en el que un grupo de expertos proporciona una etiqueta de clase para cada instancia de manera individual. Se propone un método para aprender redes bayesianas utilizando vectores de cuentas, que representan el número de expertos que seleccionan cada etiqueta de clase para cada instancia. Estas redes bayesianas se evalúan utilizando bases de datos artificiales de problemas de aprendizaje supervisado.

Multi-dimensional classification using Bayesian networks for stationary and evolving streaming data

Hoy en día, con la evolución continua y rápida de las tecnologías de la información y los dispositivos de computación, se recogen y almacenan continuamente grandes volúmenes de datos en distintos dominios y a través de diversas aplicaciones del mundo real. La extracción de conocimiento útil de una cantidad tan enorme de datos no se puede realizar habitualmente de forma manual, y requiere el uso de técnicas adecuadas de aprendizaje automático y de minería de datos. La clasificación es una de las técnicas más importantes que ha sido aplicada con éxito a varias áreas. En general, la clasificación se compone de dos pasos principales: en primer lugar, aprender un modelo de clasificación o clasificador a partir de un conjunto de datos de entrenamiento, y en segundo lugar, clasificar las nuevas instancias de datos utilizando el clasificador aprendido. La clasificación es supervisada cuando todas las etiquetas están presentes en los datos de entrenamiento (es decir, datos completamente etiquetados), semi-supervisada cuando sólo algunas etiquetas son conocidas (es decir, datos parcialmente etiquetados), y no supervisada cuando todas las etiquetas están ausentes en los datos de entrenamiento (es decir, datos no etiquetados). Además, aparte de esta taxonomía, el problema de clasificación se puede categorizar en unidimensional o multidimensional en función del número de variables clase, una o más, respectivamente; o también puede ser categorizado en estacionario o cambiante con el tiempo en función de las características de los datos y de la tasa de cambio subyacente. A lo largo de esta tesis, tratamos el problema de clasificación desde tres perspectivas diferentes, a saber, clasificación supervisada multidimensional estacionaria, clasificación semisupervisada unidimensional cambiante con el tiempo, y clasificación supervisada multidimensional cambiante con el tiempo. Para llevar a cabo esta tarea, hemos usado básicamente los clasificadores Bayesianos como modelos. La primera contribución, dirigiéndose al problema de clasificación supervisada multidimensional estacionaria, se compone de dos nuevos métodos de aprendizaje de clasificadores Bayesianos multidimensionales a partir de datos estacionarios. Los métodos se proponen desde dos puntos de vista diferentes. El primer método, denominado CB-MBC, se basa en una estrategia de envoltura de selección de variables que es voraz y hacia delante, mientras que el segundo, denominado MB-MBC, es una estrategia de filtrado de variables con una aproximación basada en restricciones y en el manto de Markov. Ambos métodos han sido aplicados a dos problemas reales importantes, a saber, la predicción de los inhibidores de la transcriptasa inversa y de la proteasa para el problema de infección por el virus de la inmunodeficiencia humana tipo 1 (HIV-1), y la predicción del European Quality of Life-5 Dimensions (EQ-5D) a partir de los cuestionarios de la enfermedad de Parkinson con 39 ítems (PDQ-39). El estudio experimental incluye comparaciones de CB-MBC y MB-MBC con los métodos del estado del arte de la clasificación multidimensional, así como con métodos comúnmente utilizados para resolver el problema de predicción de la enfermedad de Parkinson, a saber, la regresión logística multinomial, mínimos cuadrados ordinarios, y mínimas desviaciones absolutas censuradas. En ambas aplicaciones, los resultados han sido prometedores con respecto a la precisión de la clasificación, así como en relación al análisis de las estructuras gráficas que identifican interacciones conocidas y novedosas entre las variables. La segunda contribución, referida al problema de clasificación semi-supervisada unidimensional cambiante con el tiempo, consiste en un método nuevo (CPL-DS) para clasificar flujos de datos parcialmente etiquetados. Los flujos de datos difieren de los conjuntos de datos estacionarios en su proceso de generación muy rápido y en su aspecto de cambio de concepto. Es decir, los conceptos aprendidos y/o la distribución subyacente están probablemente cambiando y evolucionando en el tiempo, lo que hace que el modelo de clasificación actual sea obsoleto y deba ser actualizado. CPL-DS utiliza la divergencia de Kullback-Leibler y el método de bootstrapping para cuantificar y detectar tres tipos posibles de cambio: en las predictoras, en la a posteriori de la clase o en ambas. Después, si se detecta cualquier cambio, un nuevo modelo de clasificación se aprende usando el algoritmo EM; si no, el modelo de clasificación actual se mantiene sin modificaciones. CPL-DS es general, ya que puede ser aplicado a varios modelos de clasificación. Usando dos modelos diferentes, el clasificador naive Bayes y la regresión logística, CPL-DS se ha probado con flujos de datos sintéticos y también se ha aplicado al problema real de la detección de código malware, en el cual los nuevos ficheros recibidos deben ser continuamente clasificados en malware o goodware. Los resultados experimentales muestran que nuestro método es efectivo para la detección de diferentes tipos de cambio a partir de los flujos de datos parcialmente etiquetados y también tiene una buena precisión de la clasificación. Finalmente, la tercera contribución, sobre el problema de clasificación supervisada multidimensional cambiante con el tiempo, consiste en dos métodos adaptativos, a saber, Locally Adpative-MB-MBC (LA-MB-MBC) y Globally Adpative-MB-MBC (GA-MB-MBC). Ambos métodos monitorizan el cambio de concepto a lo largo del tiempo utilizando la log-verosimilitud media como métrica y el test de Page-Hinkley. Luego, si se detecta un cambio de concepto, LA-MB-MBC adapta el actual clasificador Bayesiano multidimensional localmente alrededor de cada nodo cambiado, mientras que GA-MB-MBC aprende un nuevo clasificador Bayesiano multidimensional. El estudio experimental realizado usando flujos de datos sintéticos multidimensionales indica los méritos de los métodos adaptativos propuestos. ABSTRACT Nowadays, with the ongoing and rapid evolution of information technology and computing devices, large volumes of data are continuously collected and stored in different domains and through various real-world applications. Extracting useful knowledge from such a huge amount of data usually cannot be performed manually, and requires the use of adequate machine learning and data mining techniques. Classification is one of the most important techniques that has been successfully applied to several areas. Roughly speaking, classification consists of two main steps: first, learn a classification model or classifier from an available training data, and secondly, classify the new incoming unseen data instances using the learned classifier. Classification is supervised when the whole class values are present in the training data (i.e., fully labeled data), semi-supervised when only some class values are known (i.e., partially labeled data), and unsupervised when the whole class values are missing in the training data (i.e., unlabeled data). In addition, besides this taxonomy, the classification problem can be categorized into uni-dimensional or multi-dimensional depending on the number of class variables, one or more, respectively; or can be also categorized into stationary or streaming depending on the characteristics of the data and the rate of change underlying it. Through this thesis, we deal with the classification problem under three different settings, namely, supervised multi-dimensional stationary classification, semi-supervised unidimensional streaming classification, and supervised multi-dimensional streaming classification. To accomplish this task, we basically used Bayesian network classifiers as models. The first contribution, addressing the supervised multi-dimensional stationary classification problem, consists of two new methods for learning multi-dimensional Bayesian network classifiers from stationary data. They are proposed from two different points of view. The first method, named CB-MBC, is based on a wrapper greedy forward selection approach, while the second one, named MB-MBC, is a filter constraint-based approach based on Markov blankets. Both methods are applied to two important real-world problems, namely, the prediction of the human immunodeficiency virus type 1 (HIV-1) reverse transcriptase and protease inhibitors, and the prediction of the European Quality of Life-5 Dimensions (EQ-5D) from 39-item Parkinson’s Disease Questionnaire (PDQ-39). The experimental study includes comparisons of CB-MBC and MB-MBC against state-of-the-art multi-dimensional classification methods, as well as against commonly used methods for solving the Parkinson’s disease prediction problem, namely, multinomial logistic regression, ordinary least squares, and censored least absolute deviations. For both considered case studies, results are promising in terms of classification accuracy as well as regarding the analysis of the learned MBC graphical structures identifying known and novel interactions among variables. The second contribution, addressing the semi-supervised uni-dimensional streaming classification problem, consists of a novel method (CPL-DS) for classifying partially labeled data streams. Data streams differ from the stationary data sets by their highly rapid generation process and their concept-drifting aspect. That is, the learned concepts and/or the underlying distribution are likely changing and evolving over time, which makes the current classification model out-of-date requiring to be updated. CPL-DS uses the Kullback-Leibler divergence and bootstrapping method to quantify and detect three possible kinds of drift: feature, conditional or dual. Then, if any occurs, a new classification model is learned using the expectation-maximization algorithm; otherwise, the current classification model is kept unchanged. CPL-DS is general as it can be applied to several classification models. Using two different models, namely, naive Bayes classifier and logistic regression, CPL-DS is tested with synthetic data streams and applied to the real-world problem of malware detection, where the new received files should be continuously classified into malware or goodware. Experimental results show that our approach is effective for detecting different kinds of drift from partially labeled data streams, as well as having a good classification performance. Finally, the third contribution, addressing the supervised multi-dimensional streaming classification problem, consists of two adaptive methods, namely, Locally Adaptive-MB-MBC (LA-MB-MBC) and Globally Adaptive-MB-MBC (GA-MB-MBC). Both methods monitor the concept drift over time using the average log-likelihood score and the Page-Hinkley test. Then, if a drift is detected, LA-MB-MBC adapts the current multi-dimensional Bayesian network classifier locally around each changed node, whereas GA-MB-MBC learns a new multi-dimensional Bayesian network classifier from scratch. Experimental study carried out using synthetic multi-dimensional data streams shows the merits of both proposed adaptive methods.

Design, implementation and evaluation of an unconstrained and contactless biometric system based on hand geometry and stress detection

Esta tesis propone un sistema biométrico de geometría de mano orientado a entornos sin contacto junto con un sistema de detección de estrés capaz de decir qué grado de estrés tiene una determinada persona en base a señales fisiológicas Con respecto al sistema biométrico, esta tesis contribuye con el diseño y la implementación de un sistema biométrico de geometría de mano, donde la adquisición se realiza sin ningún tipo de contacto, y el patrón del usuario se crea considerando únicamente datos del propio individuo. Además, esta tesis propone un algoritmo de segmentación multiescala para solucionar los problemas que conlleva la adquisición de manos en entornos reales. Por otro lado, respecto a la extracción de características y su posterior comparación esta tesis tiene una contribución específica, proponiendo esquemas adecuados para llevar a cabo tales tareas con un coste computacional bajo pero con una alta precisión en el reconocimiento de personas. Por último, este sistema es evaluado acorde a la norma estándar ISO/IEC 19795 considerando seis bases de datos públicas. En relación al método de detección de estrés, esta tesis propone un sistema basado en dos señales fisiológicas, concretamente la tasa cardiaca y la conductancia de la piel, así como la creación de un innovador patrón de estrés que recoge el comportamiento de ambas señales bajo las situaciones de estrés y no-estrés. Además, este sistema está basado en lógica difusa para decidir el grado de estrés de un individuo. En general, este sistema es capaz de detectar estrés de forma precisa y en tiempo real, proporcionando una solución adecuada para sistemas biométricos actuales, donde la aplicación del sistema de detección de estrés es directa para evitar situaciónes donde los individuos sean forzados a proporcionar sus datos biométricos. Finalmente, esta tesis incluye un estudio de aceptabilidad del usuario, donde se evalúa cuál es la aceptación del usuario con respecto a la técnica biométrica propuesta por un total de 250 usuarios. Además se incluye un prototipo implementado en un dispositivo móvil y su evaluación. ABSTRACT: This thesis proposes a hand biometric system oriented to unconstrained and contactless scenarios together with a stress detection method able to elucidate to what extent an individual is under stress based on physiological signals. Concerning the biometric system, this thesis contributes with the design and implementation of a hand-based biometric system, where the acquisition is carried out without contact and the template is created only requiring information from a single individual. In addition, this thesis proposes an algorithm based on multiscale aggregation in order to tackle with the problem of segmentation in real unconstrained environments. Furthermore, feature extraction and matching are also a specific contributions of this thesis, providing adequate schemes to carry out both actions with low computational cost but with certain recognition accuracy. Finally, this system is evaluated according to international standard ISO/IEC 19795 considering six public databases. In relation to the stress detection method, this thesis proposes a system based on two physiological signals, namely heart rate and galvanic skin response, with the creation of an innovative stress detection template which gathers the behaviour of both physiological signals under both stressing and non-stressing situations. Besides, this system is based on fuzzy logic to elucidate the level of stress of an individual. As an overview, this system is able to detect stress accurately and in real-time, providing an adequate solution for current biometric systems, where the application of a stress detection system is direct to avoid situations where individuals are forced to provide the biometric data. Finally, this thesis includes a user acceptability evaluation, where the acceptance of the proposed biometric technique is assessed by a total of 250 individuals. In addition, this thesis includes a mobile implementation prototype and its evaluation.

Automatic segmentation of abdominal aortic aneurysm from medical images based on active shape models and texture models

A semi-automatic segmentation algorithm for abdominal aortic aneurysms (AAA), and based on Active Shape Models (ASM) and texture models, is presented in this work. The texture information is provided by a set of four 3D magnetic resonance (MR) images, composed of axial slices of the abdomen, where lumen, wall and intraluminal thrombus (ILT) are visible. Due to the reduced number of images in the MRI training set, an ASM and a custom texture model based on border intensity statistics are constructed. For the same reason the shape is characterized from 35-computed tomography angiography (CTA) images set so the shape variations are better represented. For the evaluation, leave-one-out experiments have been held over the four MRI set.

Optical regeneration based on noise generated in bistable devices: going from 2R to 3R

In this paper we propose to employ an instability that occurs in bistable devices as a control signal at the reception stage to generate the clock signal. One of the adopted configurations is composed of two semiconductor optical amplifiers arranged in a cascaded structure. This configuration has an output equivalent to that obtained from Self-Electrooptic Effect Devices (SEEDs), and it can implement the main Boolean functions of two binary inputs. These outputs, obtained from the addition of two binary signals, show a short spike in the transition from "1" to "2" in the internal processing. A similar result is obtained for a simple semiconductor amplifier with bistable behavior. The paper will show how these structures may help recover clock signals in any optical transmission system

Automatic 3DTV Quality Assessment Based On Depth Perception Analysis

Quality assessment is a key factor for stereoscopic 3D video content as some observers are affected by visual discomfort in the eye when viewing 3D video, especially when combining positive and negative parallax with fast motion. In this paper, we propose techniques to assess objective quality related to motion and depth maps, which facilitate depth perception analysis. Subjective tests were carried out in order to understand the source of the problem. Motion is an important feature affecting 3D experience but also often the cause of visual discomfort. The automatic algorithm developed tries to quantify the impact on viewer experience when common cases of discomfort occur, such as high-motion sequences, scene changes with abrupt parallax changes, or complete absence of stereoscopy, with a goal of preventing the viewer from having a bad stereoscopic experience.

Managing drought economic effects through insurance schemes based on local water availability.

In many arid or semi-arid Mediterranean regions, agriculture is dependent on irrigation. When hydrological drought phenomena occur, farmers suffer from water shortages, incurring important economic losses. Yet, there is not agricultural insurance available for lack of irrigation water. This work attempts to evaluate hydrological drought risk and its economic impact on crop production in order to provide the basis for the design of drought insurance for irrigated arable crops. With this objective a model that relates water availability with expected yields is developed. Crop water requirements are calculated from evapotranspiration, effective rainfall and soil water balance. FAO?s methodology and AquaCrop software have been used to establish the relationship between water allocations and crop yields. The analysis is applied to the irrigation zone ?Riegos de Bardenas?, which is located in the Ebro river basin, northeast Spain, to the main arable crops in the area. Results show the fair premiums of different hydrological drought insurance products. Whole-farm insurance or irrigation district insurance should be preferable to crop specific insurance due to the drought management strategies used by farmers.

Neuroanatomic-based detection algorithm for automatic labeling of brain structures in brain injury

The number and grade of injured neuroanatomic structures and the type of injury determine the degree of impairment after a brain injury event and the recovery options of the patient. However, the body of knowledge and clinical intervention guides are basically focused on functional disorder and they still do not take into account the location of injuries. The prognostic value of location information is not known in detail either. This paper proposes a feature-based detection algorithm, named Neuroanatomic-Based Detection Algorithm (NBDA), based on SURF (Speeded Up Robust Feature) to label anatomical brain structures on cortical and sub-cortical areas. Themain goal is to register injured neuroanatomic structures to generate a database containing patient?s structural impairment profile. This kind of information permits to establish a relation with functional disorders and the prognostic evolution during neurorehabilitation procedures.

New generation of human machine interfaces for controlling UAV through depth based gesture recognition

New forms of natural interactions between human operators and UAVs (Unmanned Aerial Vehicle) are demanded by the military industry to achieve a better balance of the UAV control and the burden of the human operator. In this work, a human machine interface (HMI) based on a novel gesture recognition system using depth imagery is proposed for the control of UAVs. Hand gesture recognition based on depth imagery is a promising approach for HMIs because it is more intuitive, natural, and non-intrusive than other alternatives using complex controllers. The proposed system is based on a Support Vector Machine (SVM) classifier that uses spatio-temporal depth descriptors as input features. The designed descriptor is based on a variation of the Local Binary Pattern (LBP) technique to efficiently work with depth video sequences. Other major consideration is the especial hand sign language used for the UAV control. A tradeoff between the use of natural hand signs and the minimization of the inter-sign interference has been established. Promising results have been achieved in a depth based database of hand gestures especially developed for the validation of the proposed system.

A scalable approach for content based image retrieval in cloud datacenter

The emergence of cloud datacenters enhances the capability of online data storage. Since massive data is stored in datacenters, it is necessary to effectively locate and access interest data in such a distributed system. However, traditional search techniques only allow users to search images over exact-match keywords through a centralized index. These techniques cannot satisfy the requirements of content based image retrieval (CBIR). In this paper, we propose a scalable image retrieval framework which can efficiently support content similarity search and semantic search in the distributed environment. Its key idea is to integrate image feature vectors into distributed hash tables (DHTs) by exploiting the property of locality sensitive hashing (LSH). Thus, images with similar content are most likely gathered into the same node without the knowledge of any global information. For searching semantically close images, the relevance feedback is adopted in our system to overcome the gap between low-level features and high-level features. We show that our approach yields high recall rate with good load balance and only requires a few number of hops.

Integrated assessment of policy interventions for promoting sustainable irrigation in semi-arid environments: A hydro-economic modeling approach

Sustaining irrigated agriculture to meet food production needs while maintaining aquatic ecosystems is at the heart of many policy debates in various parts of the world, especially in arid and semi-arid areas. Researchers and practitioners are increasingly calling for integrated approaches, and policy-makers are progressively supporting the inclusion of ecological and social aspects in water management programs. This paper contributes to this policy debate by providing an integrated economic-hydrologic modeling framework that captures the socio-economic and environmental effects of various policy initiatives and climate variability. This modeling integration includes a risk-based economic optimization model and a hydrologic water management simulation model that have been specified for the Middle Guadiana basin, a vulnerable drought-prone agro-ecological area with highly regulated river systems in southwest Spain. Namely, two key water policy interventions were investigated: the implementation of minimum environmental flows (supported by the European Water Framework Directive, EU WFD), and a reduction in the legal amount of water delivered for irrigation (planned measure included in the new Guadiana River Basin Management Plan, GRBMP, still under discussion). Results indicate that current patterns of excessive water use for irrigation in the basin may put environmental flow demands at risk, jeopardizing the WFD s goal of restoring the ?good ecological status? of water bodies by 2015. Conflicts between environmental and agricultural water uses will be stressed during prolonged dry episodes, and particularly in summer low-flow periods, when there is an important increase of crop irrigation water requirements. Securing minimum stream flows would entail a substantial reduction in irrigation water use for rice cultivation, which might affect the profitability and economic viability of small rice-growing farms located upstream in the river. The new GRBMP could contribute to balance competing water demands in the basin and to increase economic water productivity, but might not be sufficient to ensure the provision of environmental flows as required by the WFD. A thoroughly revision of the basin s water use concession system for irrigation seems to be needed in order to bring the GRBMP in line with the WFD objectives. Furthermore, the study illustrates that social, economic, institutional, and technological factors, in addition to bio-physical conditions, are important issues to be considered for designing and developing water management strategies. The research initiative presented in this paper demonstrates that hydro-economic models can explicitly integrate all these issues, constituting a valuable tool that could assist policy makers for implementing sustainable irrigation policies.

A Hierarchical Tracking Strategy for Vision-Based Applications On-Board UAVs

In this paper, we apply a hierarchical tracking strategy of planar objects (or that can be assumed to be planar) that is based on direct methods for vision-based applications on-board UAVs. The use of this tracking strategy allows to achieve the tasks at real-time frame rates and to overcome problems posed by the challenging conditions of the tasks: e.g. constant vibrations, fast 3D changes, or limited capacity on-board. The vast majority of approaches make use of feature-based methods to track objects. Nonetheless, in this paper we show that although some of these feature-based solutions are faster, direct methods can be more robust under fast 3D motions (fast changes in position), some changes in appearance, constant vibrations (without requiring any specific hardware or software for video stabilization), and situations in which part of the object to track is outside of the field of view of the camera. The performance of the proposed tracking strategy on-board UAVs is evaluated with images from realflight tests using manually-generated ground truth information, accurate position estimation using a Vicon system, and also with simulated data from a simulation environment. Results show that the hierarchical tracking strategy performs better than wellknown feature-based algorithms and well-known configurations of direct methods, and that its performance is robust enough for vision-in-the-loop tasks, e.g. for vision-based landing tasks.

Effect of conservation tillage on soil physical properties, in-situ rainwater harvesting, and erosion control in arid and semi-arid regions

La escasez del agua en las regiones áridas y semiáridas se debe a la escasez de precipitaciones y la distribución desigual en toda la temporada, lo que hace de la agricultura de secano una empresa precaria. Un enfoque para mejorar y estabilizar el agua disponible para la producción de cultivos en estas regiones es el uso de tecnologías de captación de agua de lluvia in situ y su conservación. La adopción de los sistemas de conservación de la humedad del suelo in situ, tales como la labranza de conservación, es una de las estrategias para mejorar la gestión de la agricultura en zonas áridas y semiáridas. El objetivo general de esta tesis ha sido desarrollar una metodología de aplicación de labranza de depósito e investigar los efectos a corto plazo sobre las propiedades físicas del suelo de las diferentes prácticas de cultivo que incluyen labranza de depósito: (reservoir tillage, RT), la laboreo mínimo: (minimum tillage, MT), la no laboreo: (zero tillage, ZT) y laboreo convencional: (conventional tillage, CT) Así como, la retención de agua del suelo y el control de la erosión del suelo en las zonas áridas y semiáridas. Como una primera aproximación, se ha realizado una revisión profunda del estado de la técnica, después de la cual, se encontró que la labranza de depósito es un sistema eficaz de cosecha del agua de lluvia y conservación del suelo, pero que no ha sido evaluada científicamente tanto como otros sistemas de labranza. Los trabajos experimentales cubrieron tres condiciones diferentes: experimentos en laboratorio, experimentos de campo en una región árida, y experimentos de campo en una región semiárida. Para investigar y cuantificar el almacenamiento de agua a temperatura ambiente y la forma en que podría adaptarse para mejorar la infiltración del agua de lluvia recolectada y reducir la erosión del suelo, se ha desarrollado un simulador de lluvia a escala de laboratorio. Las características de las lluvias, entre ellas la intensidad de las precipitaciones, la uniformidad espacial y tamaño de la gota de lluvia, confirmaron que las condiciones naturales de precipitación son simuladas con suficiente precisión. El simulador fue controlado automáticamente mediante una válvula de solenoide y tres boquillas de presión que se usaron para rociar agua correspondiente a diferentes intensidades de lluvia. Con el fin de evaluar el método de RT bajo diferentes pendientes de superficie, se utilizaron diferentes dispositivos de pala de suelo para sacar un volumen idéntico para hacer depresiones. Estas depresiones se compararon con una superficie de suelo control sin depresión, y los resultados mostraron que la RT fue capaz de reducir la erosión del suelo y la escorrentía superficial y aumentar significativamente la infiltración. Luego, basándonos en estos resultados, y después de identificar la forma adecuada de las depresiones, se ha diseñado una herramienta combinada (sistema integrado de labranza de depósito (RT)) compuesto por un arado de una sola línea de chisel, una sola línea de grada en diente de pico, sembradora modificada, y rodillo de púas. El equipo fue construido y se utiliza para comparación con MT y CT en un ambiente árido en Egipto. El estudio se realizó para evaluar el impacto de diferentes prácticas de labranza y sus parámetros de funcionamiento a diferentes profundidades de labranza y con distintas velocidades de avance sobre las propiedades físicas del suelo, así como, la pérdida de suelo, régimen de humedad, la eficiencia de recolección de agua, y la productividad de trigo de invierno. Los resultados indicaron que la RT aumentó drásticamente la infiltración, produciendo una tasa que era 47.51% más alta que MT y 64.56% mayor que la CT. Además, los resultados mostraron que los valores más bajos de la escorrentía y pérdidas de suelos 4.91 mm y 0.65 t ha-1, respectivamente, se registraron en la RT, mientras que los valores más altos, 11.36 mm y 1.66 t ha-1, respectivamente, se produjeron en el marco del CT. Además, otros dos experimentos de campo se llevaron a cabo en ambiente semiárido en Madrid con la cebada y el maíz como los principales cultivos. También ha sido estudiado el potencial de la tecnología inalámbrica de sensores para monitorizar el potencial de agua del suelo. Para el experimento en el que se cultivaba la cebada en secano, se realizaron dos prácticas de labranza (RT y MT). Los resultados mostraron que el potencial del agua del suelo aumentó de forma constante y fue consistentemente mayor en MT. Además, con independencia de todo el período de observación, RT redujo el potencial hídrico del suelo en un 43.6, 5.7 y 82.3% respectivamente en comparación con el MT a profundidades de suelo (10, 20 y 30 cm, respectivamente). También se observaron diferencias claras en los componentes del rendimiento de los cultivos y de rendimiento entre los dos sistemas de labranza, el rendimiento de grano (hasta 14%) y la producción de biomasa (hasta 8.8%) se incrementaron en RT. En el experimento donde se cultivó el maíz en regadío, se realizaron cuatro prácticas de labranza (RT, MT, ZT y CT). Los resultados revelaron que ZT y RT tenían el potencial de agua y temperatura del suelo más bajas. En comparación con el tratamiento con CT, ZT y RT disminuyó el potencial hídrico del suelo en un 72 y 23%, respectivamente, a la profundidad del suelo de 40 cm, y provocó la disminución de la temperatura del suelo en 1.1 y un 0.8 0C respectivamente, en la profundidad del suelo de 5 cm y, por otro lado, el ZT tenía la densidad aparente del suelo y resistencia a la penetración más altas, la cual retrasó el crecimiento del maíz y disminuyó el rendimiento de grano que fue del 15.4% menor que el tratamiento con CT. RT aumenta el rendimiento de grano de maíz cerca de 12.8% en comparación con la ZT. Por otra parte, no hubo diferencias significativas entre (RT, MT y CT) sobre el rendimiento del maíz. En resumen, según los resultados de estos experimentos, se puede decir que mediante el uso de la labranza de depósito, consistente en realizar depresiones después de la siembra, las superficies internas de estas depresiones se consolidan de tal manera que el agua se mantiene para filtrarse en el suelo y por lo tanto dan tiempo para aportar humedad a la zona de enraizamiento de las plantas durante un período prolongado de tiempo. La labranza del depósito podría ser utilizada como un método alternativo en regiones áridas y semiáridas dado que retiene la humedad in situ, a través de estructuras que reducen la escorrentía y por lo tanto puede resultar en la mejora de rendimiento de los cultivos. ABSTRACT Water shortage in arid and semi-arid regions stems from low rainfall and uneven distribution throughout the season, which makes rainfed agriculture a precarious enterprise. One approach to enhance and stabilize the water available for crop production in these regions is to use in-situ rainwater harvesting and conservation technologies. Adoption of in-situ soil moisture conservation systems, such as conservation tillage, is one of the strategies for upgrading agriculture management in arid and semi-arid environments. The general aim of this thesis is to develop a methodology to apply reservoir tillage to investigate the short-term effects of different tillage practices including reservoir tillage (RT), minimum tillage (MT), zero tillage (ZT), and conventional tillage (CT) on soil physical properties, as well as, soil water retention, and soil erosion control in arid and semi-arid areas. As a first approach, a review of the state of the art has been done. We found that reservoir tillage is an effective system of harvesting rainwater and conserving soil, but it has not been scientifically evaluated like other tillage systems. Experimental works covered three different conditions: laboratory experiments, field experiments in an arid region, and field experiments in a semi-arid region. To investigate and quantify water storage from RT and how it could be adapted to improve infiltration of harvested rainwater and reduce soil erosion, a laboratory-scale rainfall simulator was developed. Rainfall characteristics, including rainfall intensity, spatial uniformity and raindrop size, confirm that natural rainfall conditions are simulated with sufficient accuracy. The simulator was auto-controlled by a solenoid valve and three pressure nozzles were used to spray water corresponding to different rainfall intensities. In order to assess the RT method under different surface slopes, different soil scooping devices with identical volume were used to create depressions. The performance of the soil with these depressions was compared to a control soil surface (with no depression). Results show that RT was able to reduce soil erosion and surface runoff and significantly increase infiltration. Then, based on these results and after selecting the proper shape of depressions, a combination implement integrated reservoir tillage system (integrated RT) comprised of a single-row chisel plow, single-row spike tooth harrow, modified seeder, and spiked roller was developed and used to compared to MT and CT in an arid environment in Egypt. The field experiments were conducted to evaluate the impact of different tillage practices and their operating parameters at different tillage depths and different forward speeds on the soil physical properties, as well as on runoff, soil losses, moisture regime, water harvesting efficiency, and winter wheat productivity. Results indicated that the integrated RT drastically increased infiltration, producing a rate that was 47.51% higher than MT and 64.56% higher than CT. In addition, results showed that the lowest values of runoff and soil losses, 4.91 mm and 0.65 t ha-1 respectively, were recorded under the integrated RT, while the highest values, 11.36 mm and 1.66 t ha -1 respectively, occurred under the CT. In addition, two field experiments were carried out in semi-arid environment in Madrid with barley and maize as the main crops. For the rainfed barley experiment, two tillage practices (RT, and MT) were performed. Results showed that soil water potential increased quite steadily and were consistently greater in MT and, irrespective of the entire observation period, RT decreased soil water potential by 43.6, 5.7, and 82.3% compared to MT at soil depths (10, 20, and 30 cm, respectively). In addition, clear differences in crop yield and yield components were observed between the two tillage systems, grain yield (up to 14%) and biomass yield (up to 8.8%) were increased by RT. For the irrigated maize experiment, four tillage practices (RT, MT, ZT, and CT) were performed. Results showed that ZT and RT had the lowest soil water potential and soil temperature. Compared to CT treatment, ZT and RT decreased soil water potential by 72 and 23% respectively, at soil depth of 40 cm, and decreased soil temperature by 1.1 and 0.8 0C respectively, at soil depth of 5 cm. Also, ZT had the highest soil bulk density and penetration resistance, which delayed the maize growth and decreased the grain yield that was 15.4% lower than CT treatment. RT increased maize grain yield about 12.8% compared to ZT. On the other hand, no significant differences among (RT, MT, and CT) on maize yield were found. In summary, according to the results from these experiments using reservoir tillage to make depressions after seeding, these depression’s internal surfaces are consolidated in such a way that the water is held to percolate into the soil and thus allowing time to offer moisture to the plant rooting zone over an extended period of time. Reservoir tillage could be used as an alternative method in arid and semi-arid regions and it retains moisture in-situ, through structures that reduce runoff and thus can result in improved crop yields.

HOG-Like gradient-based descriptor for visual vehicle detection

One of the main challenges for intelligent vehicles is the capability of detecting other vehicles in their environment, which constitute the main source of accidents. Specifically, many methods have been proposed in the literature for video-based vehicle detection. Most of them perform supervised classification using some appearance-related feature, in particular, symmetry has been extensively utilized. However, an in-depth analysis of the classification power of this feature is missing. As a first contribution of this paper, a thorough study of the classification performance of symmetry is presented within a Bayesian decision framework. This study reveals that the performance of symmetry-based classification is very limited. Therefore, as a second contribution, a new gradient-based descriptor is proposed for vehicle detection. This descriptor exploits the known rectangular structure of vehicle rears within a Histogram of Gradients (HOG)-based framework. Experiments show that the proposed descriptor outperforms largely symmetry as a feature for vehicle verification, achieving classification rates over 90%.

Efficient hybrid monocular-stereo approach to on-board, video-based traffic sign detection and tracking

In this paper we propose an innovative method for the automatic detection and tracking of road traffic signs using an onboard stereo camera. It involves a combination of monocular and stereo analysis strategies to increase the reliability of the detections such that it can boost the performance of any traffic sign recognition scheme. Firstly, an adaptive color and appearance based detection is applied at single camera level to generate a set of traffic sign hypotheses. In turn, stereo information allows for sparse 3D reconstruction of potential traffic signs through a SURF-based matching strategy. Namely, the plane that best fits the cloud of 3D points traced back from feature matches is estimated using a RANSAC based approach to improve robustness to outliers. Temporal consistency of the 3D information is ensured through a Kalman-based tracking stage. This also allows for the generation of a predicted 3D traffic sign model, which is in turn used to enhance the previously mentioned color-based detector through a feedback loop, thus improving detection accuracy. The proposed solution has been tested with real sequences under several illumination conditions and in both urban areas and highways, achieving very high detection rates in challenging environments, including rapid motion and significant perspective distortion