Der Renormierungsgruppen-Fluß der konform-reduzierten Quantengravitation

Wir analysieren die Rolle von "Hintergrundunabhängigkeit" im Zugang der effektiven Mittelwertwirkung zur Quantengravitation. Wenn der nicht-störungstheoretische Renormierungsgruppen-(RG)-Fluß "hintergrundunabhängig" ist, muß die Vergröberung durch eine nicht spezifizierte, variable Metrik definiert werden. Die Forderung nach "Hintergrundunabhängigkeit" in der Quantengravitation führt dazu, daß die funktionale RG-Gleichung von zusätzlichen Feldern abhängt; dadurch unterscheidet sich der RG-Fluß in der Quantengravitation deutlich von dem RG-Fluß einer gewöhnlichen Quantentheorie, deren Moden-Cutoff von einer starren Metrik abhängt. Beispielsweise kann in der "hintergrundunabhängigen" Theorie ein Nicht-Gauß'scher Fixpunkt existieren, obwohl die entsprechende gewöhnliche Quantentheorie keinen solchen entwickelt. Wir untersuchen die Bedeutung dieses universellen, rein kinematischen Effektes, indem wir den RG-Fluß der Quanten-Einstein-Gravitation (QEG) in einem "konform-reduzierten" Zusammenhang untersuchen, in dem wir nur den konformen Faktor der Metrik quantisieren. Alle anderen Freiheitsgrade der Metrik werden vernachlässigt. Die konforme Reduktion der Einstein-Hilbert-Trunkierung zeigt exakt dieselben qualitativen Eigenschaften wie in der vollen Einstein-Hilbert-Trunkierung. Insbesondere besitzt sie einen Nicht-Gauß'schen Fixpunkt, der notwendig ist, damit die Gravitation asymptotisch sicher ist. Ohne diese zusätzlichen Feldabhängigkeiten ist der RG-Fluß dieser Trunkierung der einer gewöhnlichen $phi^4$-Theorie. Die lokale Potentialnäherung für den konformen Faktor verallgemeinert den RG-Fluß in der Quantengravitation auf einen unendlich-dimensionalen Theorienraum. Auch hier finden wir sowohl einen Gauß'schen als auch einen Nicht-Gauß'schen Fixpunkt, was weitere Hinweise dafür liefert, daß die Quantengravitation asymptotisch sicher ist. Das Analogon der Metrik-Invarianten, die proportional zur dritten Potenz der Krümmung ist und die die störungstheoretische Renormierbarkeit zerstört, ist unproblematisch für die asymptotische Sicherheit der konform-reduzierten Theorie. Wir berechnen die Skalenfelder und -imensionen der beiden Fixpunkte explizit und diskutieren mögliche Einflüsse auf die Vorhersagekraft der Theorie. Da der RG-Fluß von der Topologie der zugrundeliegenden Raumzeit abhängt, diskutieren wir sowohl den flachen Raum als auch die Sphäre. Wir lösen die Flußgleichung für das Potential numerisch und erhalten Beispiele für RG-Trajektorien, die innerhalb der Ultraviolett-kritischen Mannigfaltigkeit des Nicht-Gauß'schen Fixpunktes liegen. Die Quantentheorien, die durch einige solcher Trajektorien definiert sind, zeigen einen Phasenübergang von der bekannten (Niederenergie-) Phase der Gravitation mit spontan gebrochener Diffeomorphismus-Invarianz zu einer neuen Phase von ungebrochener Diffeomorphismus-Invarianz. Diese Hochenergie-Phase ist durch einen verschwindenden Metrik-Erwartungswert charakterisiert.

Charge and spin transport in nanoscale junction from first principles

Recently nanoscale junctions consisting of 0-D nanostructures (single molecule) or 1-D nanostructures (semiconducting nanowire) sandwiched between two metal electrodes are successfully fabricated and characterized. What lacks in the recent developments is the understanding of the mechanism behind the observed phenomena at the level of atoms and electrons. For example, the origin of observed switching effect in a semiconducting nanowire due to the influence of an external gate bias is not yet understood at the electronic structure level. On the same context, different experimental groups have reported different signs in tunneling magneto-resistance for the same organic spin valve structure, which has baffled researchers working in this field. In this thesis, we present the answers to some of these subtle questions by investigating the charge and spin transport in different nanoscale junctions. A parameter-free, single particle Green’s function approach in conjunction with a posteriori density functional theory (DFT) involving a hybrid orbital dependent functional is used to calculate the tunneling current in the coherent transport limit. The effect of spin polarization is explicitly incorporated to investigate spin transport in a nanoscale junction. Through the electron transport studies in PbS nanowire junction, a new orbital controlled mechanism behind the switching of the current is proposed. It can explain the switching behavior, not only in PbS nanowire, but in other lead-chalcogenide nanowires as well. Beside this, the electronic structure properties of this nanowire are studied using periodic DFT. The quantum confinement effect was investigated by calculating the bandgap of PbS nanowires with different diameters. Subsequently, we explain an observed semiconducting to metallic phase transition of this nanowire by calculating the bandgap of the nanowire under uniform radial strain. The compressive radial strain on the nanowire was found to be responsible for the metallic to semiconducting phase transition. Apart from studying one dimensional nanostructure, we also present transport properties in zero dimensional single molecular junctions. We proposed a new codoping approach in a single molecular carborane junction, where a cation and an anion are simultaneously doped to find the role of a single atom in the device. The main purpose was to build a molecular junction where a single atom can dictate the flow of electrons in a circuit. Recent observations of both positive and negative sign in tunneling magnetoresistance (TMR) the using same organic spin-valve structure hasmystified researchers. From our spin dependent transport studies in a prototypical organic molecular tunneling device, we found that a 3% change in metal-molecule interfacial distance can alter the sign of TMR. Changing the interfacial distance by 3%, the number of participating eigenstates as well as their orbital characteristic changes for anti-parallel configuration of the magnetization at the two electrodes, leading to the sign reversal of the TMR. Apart from this, the magnetic proximity effect under applied bias is investigated quantitatively, which can be used to understand the observed unexpectedmagnetismin carbon basedmaterials when they are in close proximity with magnetic substrates.

Controlling electronic and magnetic properties of ultra narrow multilayered nanowires

Interest in the study of magnetic/non-magnetic multilayered structures took a giant leap since Grünberg and his group established that the interlayer exchange coupling (IEC) is a function of the non-magnetic spacer width. This interest was further fuelled by the discovery of the phenomenal Giant Magnetoresistance (GMR) effect. In fact, in 2007 Albert Fert and Peter Grünberg were awarded the Nobel Prize in Physics for their contribution to the discovery of GMR. GMR is the key property that is being used in the read-head of the present day computer hard drive as it requires a high sensitivity in the detection of magnetic field. The recent increase in demand for device miniaturization encouraged researchers to look for GMR in nanoscale multilayered structures. In this context, one dimensional(1-D) multilayerd nanowire structure has shown tremendous promise as a viable candidate for ultra sensitive read head sensors. In fact, the phenomenal giant magnetoresistance(GMR) effect, which is the novel feature of the currently used multilayered thin film, has already been observed in multilayered nanowire systems at ambient temperature. Geometrical confinement of the supper lattice along the 2-dimensions (2-D) to construct the 1-D multilayered nanowire prohibits the minimization of magnetic interaction- offering a rich variety of magnetic properties in nanowire that can be exploited for novel functionality. In addition, introduction of non-magnetic spacer between the magnetic layers presents additional advantage in controlling magnetic properties via tuning the interlayer magnetic interaction. Despite of a large volume of theoretical works devoted towards the understanding of GMR and IEC in super lattice structures, limited theoretical calculations are reported in 1-D multilayered systems. Thus to gauge their potential application in new generation magneto-electronic devices, in this thesis, I have discussed the usage of first principles density functional theory (DFT) in predicting the equilibrium structure, stability as well as electronic and magnetic properties of one dimensional multilayered nanowires. Particularly, I have focused on the electronic and magnetic properties of Fe/Pt multilayered nanowire structures and the role of non-magnetic Pt spacer in modulating the magnetic properties of the wire. It is found that the average magnetic moment per atom in the nanowire increases monotonically with an ~1/(N(Fe)) dependance, where N(Fe) is the number of iron layers in the nanowire. A simple model based upon the interfacial structure is given to explain the 1/(N(Fe)) trend in magnetic moment obtained from the first principle calculations. A new mechanism, based upon spin flip with in the layer and multistep electron transfer between the layers, is proposed to elucidate the enhancement of magnetic moment of Iron atom at the Platinum interface. The calculated IEC in the Fe/Pt multilayered nanowire is found to switch sign as the width of the non-magnetic spacer varies. The competition among short and long range direct exchange and the super exchange has been found to play a key role for the non-monotonous sign in IEC depending upon the width of the Platinum spacer layer. The calculated magnetoresistance from Julliere's model also exhibit similar switching behavior as that of IEC. The universality of the behavior of exchange coupling has also been looked into by introducing different non-magnetic spacers like Palladium, Copper, Silver, and Gold in between magnetic Iron layers. The nature of hybridization between Fe and other non-magnetic spacer is found to dictate the inter layer magnetic interaction. For example, in Fe/Pd nanowire the d-p hybridization in two spacer layer case favors anti-ferromagnetic (AFM) configuration over ferromagnetic (FM) configuration. However, the hybridization between half-filled Fe(d) and filled Cu(p) state in Fe/Cu nanowire favors FM coupling in the 2-spacer system.

Metal-oxide film and photonic structures for integrated device applications

The application of photonic crystal technology on metal-oxide film is a very promising field for future optical telecommunication systems. Band gap and polarization effects in lithium niobate (LiNbO3) photonic crystals and bismuth-substituted iron garnets (BiYIG) photonic crystals are investigated in this work reported here. The design and fabrication process are similar for these two materials while the applications are different, involving Bragg filtering in lithium niobate and polarization rotation in nonreciprocal iron garnets. The research of photonic structures in LiNbO3 is of high interest for integrated device application due to its remarkable electro-optical characteristics. This work investigated the photonic band gap in high quality LiNbO3 single crystalline thin film by ion implantation to realize high efficiency narrow bandwidth filters. LiNbO3 thin film detachment by bonding is also demonstrated for optical device integration. One-dimensional Bragg BiYIG waveguides in gyrotropic system are found to have multiple stopbands and evince enhancement of polarization rotation efficiency. Previous photon trapping theory cannot explain the phenomena because of the presence of linear birefringence. This work is aimed at investigating the mechanism with the support of experiments. The results we obtained show that selective suppression of Bloch states in gyrotropic bandgaps is the key mechanism for the observed phenomena. Finally, the research of ferroelectric single crystal PMN-PT with ultra high piezoelectric coefficient as a biosensor is also reported. This work presents an investigation and results on higher sensitivity effects than conventional materials such as quartz and lithium niobate.

STOCHASTIC CHARGE TRANSPORT IN MULTI-ISLAND SINGLE-ELECTRON TUNNELING DEVICES

The physics of the operation of singe-electron tunneling devices (SEDs) and singe-electron tunneling transistors (SETs), especially of those with multiple nanometer-sized islands, has remained poorly understood in spite of some intensive experimental and theoretical research. This computational study examines the current-voltage (IV) characteristics of multi-island single-electron devices using a newly developed multi-island transport simulator (MITS) that is based on semi-classical tunneling theory and kinetic Monte Carlo simulation. The dependence of device characteristics on physical device parameters is explored, and the physical mechanisms that lead to the Coulomb blockade (CB) and Coulomb staircase (CS) characteristics are proposed. Simulations using MITS demonstrate that the overall IV characteristics in a device with a random distribution of islands are a result of a complex interplay among those factors that affect the tunneling rates that are fixed a priori (e.g. island sizes, island separations, temperature, gate bias, etc.), and the evolving charge state of the system, which changes as the source-drain bias (VSD) is changed. With increasing VSD, a multi-island device has to overcome multiple discrete energy barriers (up-steps) before it reaches the threshold voltage (Vth). Beyond Vth, current flow is rate-limited by slow junctions, which leads to the CS structures in the IV characteristic. Each step in the CS is characterized by a unique distribution of island charges with an associated distribution of tunneling probabilities. MITS simulation studies done on one-dimensional (1D) disordered chains show that longer chains are better suited for switching applications as Vth increases with increasing chain length. They are also able to retain CS structures at higher temperatures better than shorter chains. In sufficiently disordered 2D systems, we demonstrate that there may exist a dominant conducting path (DCP) for conduction, which makes the 2D device behave as a quasi-1D device. The existence of a DCP is sensitive to the device structure, but is robust with respect to changes in temperature, gate bias, and VSD. A side gate in 1D and 2D systems can effectively control Vth. We argue that devices with smaller island sizes and narrower junctions may be better suited for practical applications, especially at room temperature.

What is the Relation between Human Practical Action and an Accompanying Discourse? Discussing the Status of Practical Theory

The article discusses the function of an accompanying discourse in relation to the genesis of human practical action. On the one side, theory cannot be taken as the ground for practical action; practical action is not a realisation of intentions. On the other hand, human practical action is accompanied by series of explanations, justifications, declarations of intent, pre‑ and post-rationalisations, motivations etc. These accompanying discourses seem in one way or the other to be necessary for the actual realisation of human practical action. Following Pierre Bourdieu, it is suggested that an accompanying discourse cannot in a meaningful manner be separated from the human practical action, that practical theory should be regarded not as theory but as part of practice, and that practical theory first of all provides a common language for talking about practice and hence for reproducing a fundamentally arbitrary idea of the genesis of human practical action. Parallels are drawn to the education/formal training of semi-professionals.

THE INTERACTION BETWEEN INTERNAL AND EXTERNAL INFORMATION ON RELATIONAL DATA SEARCH

People often use tools to search for information. In order to improve the quality of an information search, it is important to understand how internal information, which is stored in user’s mind, and external information, represented by the interface of tools interact with each other. How information is distributed between internal and external representations significantly affects information search performance. However, few studies have examined the relationship between types of interface and types of search task in the context of information search. For a distributed information search task, how data are distributed, represented, and formatted significantly affects the user search performance in terms of response time and accuracy. Guided by UFuRT (User, Function, Representation, Task), a human-centered process, I propose a search model, task taxonomy. The model defines its relationship with other existing information models. The taxonomy clarifies the legitimate operations for each type of search task of relation data. Based on the model and taxonomy, I have also developed prototypes of interface for the search tasks of relational data. These prototypes were used for experiments. The experiments described in this study are of a within-subject design with a sample of 24 participants recruited from the graduate schools located in the Texas Medical Center. Participants performed one-dimensional nominal search tasks over nominal, ordinal, and ratio displays, and searched one-dimensional nominal, ordinal, interval, and ratio tasks over table and graph displays. Participants also performed the same task and display combination for twodimensional searches. Distributed cognition theory has been adopted as a theoretical framework for analyzing and predicting the search performance of relational data. It has been shown that the representation dimensions and data scales, as well as the search task types, are main factors in determining search efficiency and effectiveness. In particular, the more external representations used, the better search task performance, and the results suggest the ideal search performance occurs when the question type and corresponding data scale representation match. The implications of the study lie in contributing to the effective design of search interface for relational data, especially laboratory results, which are often used in healthcare activities.

Numerical study of Anderson localization of terahertz waves in disordered waveguides

We present a numerical study of electromagnetic wave transport in disordered quasi-one-dimensional waveguides at terahertz frequencies. Finite element method calculations of terahertz wave propagation within LiNbO3 waveguides with randomly arranged air-filled circular scatterers exhibit an onset of Anderson localization at experimentally accessible length scales. Results for the average transmission as a function of waveguide length and scatterer density demonstrate a clear crossover from diffusive to localized transport regime. In addition, we find that transmission fluctuations grow dramatically when crossing into the localized regime. Our numerical results are in good quantitative agreement with theory over a wide range of experimentally accessible parameters both in the diffusive and localized regime opening the path towards experimental observation of terahertz wave localization.

On the similarity of Sturm-Liouville operators with non-Hermitian boundary conditions to self-adjoint and normal operators

We consider one-dimensional Schrödinger-type operators in a bounded interval with non-self-adjoint Robin-type boundary conditions. It is well known that such operators are generically conjugate to normal operators via a similarity transformation. Motivated by recent interests in quasi-Hermitian Hamiltonians in quantum mechanics, we study properties of the transformations and similar operators in detail. In the case of parity and time reversal boundary conditions, we establish closed integral-type formulae for the similarity transformations, derive a non-local self-adjoint operator similar to the Schrödinger operator and also find the associated “charge conjugation” operator, which plays the role of fundamental symmetry in a Krein-space reformulation of the problem.

Low-dimensional modeling of streaks in a wedge flow boundary layer

This paper is concerned with the low dimensional structure of optimal streaks in a wedge flow boundary layer, which have been recently shown to consist of a unique (up to a constant factor) three-dimensional streamwise evolving mode, known as the most unstable streaky mode. Optimal streaks exhibit a still unexplored/unexploited approximate self-similarity (not associated with the boundary layer self-similarity), namely the streamwise velocity re-scaled with their maximum remains almost independent of both the spanwise wavenumber and the streamwise coordinate; the remaining two velocity components instead do not satisfy this property. The approximate self-similar behavior is analyzed here and exploited to further simplify the description of optimal streaks. In particular, it is shown that streaks can be approximately described in terms of the streamwise evolution of the scalar amplitudes of just three one-dimensional modes, providing the wall normal profiles of the streamwise velocity and two combinations of the cross flow velocity components; the scalar amplitudes obey a singular system of three ordinary differential equations (involving only two degrees of freedom), which approximates well the streamwise evolution of the general streaks.

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.

Numerical and experimental study of the dynamics of axisymmetric slender liquid bridges

A one-dimensional inviscid slice model has been used to study numerically the influence of axial microgravity on the breaking of liquid bridges having a volume close to that of gravitationless minimum volume stability limit. Equilibrium shapes and stability limits have been obtained as well as the dependence of the volume of the two drops formed after breaking on both the length and the volume of the liquid bridge. The breaking process has also been studied experimentally. Good agreement has been found between theory and experiment for neutrally buoyant systems

Systematic derivation of partition functions for ligand binding to two-dimensional lattices

The Ising problem consists in finding the analytical solution of the partition function of a lattice once the interaction geometry among its elements is specified. No general analytical solution is available for this problem, except for the one-dimensional case. Using site-specific thermodynamics, it is shown that the partition function for ligand binding to a two-dimensional lattice can be obtained from those of one-dimensional lattices with known solution. The complexity of the lattice is reduced recursively by application of a contact transformation that involves a relatively small number of steps. The transformation implemented in a computer code solves the partition function of the lattice by operating on the connectivity matrix of the graph associated with it. This provides a powerful new approach to the Ising problem, and enables a systematic analysis of two-dimensional lattices that model many biologically relevant phenomena. Application of this approach to finite two-dimensional lattices with positive cooperativity indicates that the binding capacity per site diverges as Na (N = number of sites in the lattice) and experiences a phase-transition-like discontinuity in the thermodynamic limit N → ∞. The zeroes of the partition function tend to distribute on a slightly distorted unit circle in complex plane and approach the positive real axis already for a 5×5 square lattice. When the lattice has negative cooperativity, its properties mimic those of a system composed of two classes of independent sites with the apparent population of low-affinity binding sites increasing with the size of the lattice, thereby accounting for a phenomenon encountered in many ligand-receptor interactions.

Spectral bifurcations in dispersive wave turbulence

Dispersive wave turbulence is studied numerically for a class of one-dimensional nonlinear wave equations. Both deterministic and random (white noise in time) forcings are studied. Four distinct stable spectra are observed—the direct and inverse cascades of weak turbulence (WT) theory, thermal equilibrium, and a fourth spectrum (MMT; Majda, McLaughlin, Tabak). Each spectrum can describe long-time behavior, and each can be only metastable (with quite diverse lifetimes)—depending on details of nonlinearity, forcing, and dissipation. Cases of a long-live MMT transient state dcaying to a state with WT spectra, and vice-versa, are displayed. In the case of freely decaying turbulence, without forcing, both cascades of weak turbulence are observed. These WT states constitute the clearest and most striking numerical observations of WT spectra to date—over four decades of energy, and three decades of spatial, scales. Numerical experiments that study details of the composition, coexistence, and transition between spectra are then discussed, including: (i) for deterministic forcing, sharp distinctions between focusing and defocusing nonlinearities, including the role of long wavelength instabilities, localized coherent structures, and chaotic behavior; (ii) the role of energy growth in time to monitor the selection of MMT or WT spectra; (iii) a second manifestation of the MMT spectrum as it describes a self-similar evolution of the wave, without temporal averaging; (iv) coherent structures and the evolution of the direct and inverse cascades; and (v) nonlocality (in k-space) in the transferral process.

Reflection of structural waves at a solid/liquid interface

This paper investigates the reflection characteristics of structural or guided waves in rods at a solid/liquid interface. Structural waves, whose wavelengths are much larger than the diameter of the rod, are described in a first approximation by classical one-dimensional wave theory. The reflection characteristics of such waves at a solid/liquid (melting) interface has been reported by two different ultrasonic measurement techniques: first, measuring the fast regression rate of a melting interface during the burning of metal rod samples in an oxygen-enriched environment, and second, monitoring the propagation of the solid/liquid interface during the slow melting and solidification of a rod sample in a furnace. The second work clearly shows that the major reflection occurs from the solid/liquid interface and not the liquid/gas interface as predicted by plane longitudinal wave reflectivity theory. The present work confirms this observation by reporting on the results of some specially designed experiments to identify the main interface of reflection for structural waves in rods. Hence, it helps in explaining the fundamental discrepancy between the reflection characteristics at a solid/liquid interface between low frequency structural waves and high frequency bulk waves, and confirms that the detected echo within a burning metallic rod clearly represents a reflection from the solid/liquid interface. (C) 2003 Elsevier Science B.V. All rights reserved.