Improving Wishart Classification of Polarimetric SAR Data Using the Hopfield Neural Network Optimization Approach.

This paper proposes the optimization relaxation approach based on the analogue Hopfield Neural Network (HNN) for cluster refinement of pre-classified Polarimetric Synthetic Aperture Radar (PolSAR) image data. We consider the initial classification provided by the maximum-likelihood classifier based on the complex Wishart distribution, which is then supplied to the HNN optimization approach. The goal is to improve the classification results obtained by the Wishart approach. The classification improvement is verified by computing a cluster separability coefficient and a measure of homogeneity within the clusters. During the HNN optimization process, for each iteration and for each pixel, two consistency coefficients are computed, taking into account two types of relations between the pixel under consideration and its corresponding neighbors. Based on these coefficients and on the information coming from the pixel itself, the pixel under study is re-classified. Different experiments are carried out to verify that the proposed approach outperforms other strategies, achieving the best results in terms of separability and a trade-off with the homogeneity preserving relevant structures in the image. The performance is also measured in terms of computational central processing unit (CPU) times.

Sistemas multiespectrales e hiperespectrales para la observación del territorio : análisis y aplicación a la prospección de hidrocarburos.

La observación de la Tierra es una herramienta de gran utilidad en la actualidad para el estudio de los fenómenos que se dan en la misma. La observación se puede realizar a distintas escalas y por distintos métodos dependiendo del propósito. El actual Trabajo Final de Grado persigue exponer la observación del territorio mediante técnicas de Teledetección, o Detección Remota, y su aplicación en la exploración de hidrocarburos. Desde la Segunda Guerra Mundial el capturar imágenes aéreas de regiones de la Tierra estaba restringido a usos cartográficos en el sentido estricto. Desde aquellos tiempos, hasta ahora, ha acontecido una serie de avances científicos que permiten deducir características intrínsecas de la Tierra mediante mecanismos complejos que no apreciamos a simple vista, pero que, están configurados mediante determinados parámetros geométricos y electrónicos, que permiten generar series temporales de fenómenos físicos que se dan en la Tierra. Hoy en día se puede afirmar que el aprovechamiento del espectro electromagnético está en un punto máximo. Se ha pasado del análisis de la región del espectro visible al análisis del espectro en su totalidad. Esto supone el desarrollo de nuevos algoritmos, técnicas y procesos para extraer la mayor cantidad de información acerca de la interacción de la materia con la radiación electromagnética. La información que generan los sistemas de captura va a servir para la aplicación directa e indirecta de métodos de prospección de hidrocarburos. Las técnicas utilizadas en detección por sensores remotos, aplicadas en campañas geofísicas, son utilizadas para minimizar costes y maximizar resultados en investigaciones de campo. La predicción de anomalías en la zona de estudio depende del analista, quien diseña, calcula y evalúa las variaciones de la energía electromagnética reflejada o emitida por la superficie terrestre. Para dicha predicción se revisarán distintos programas espaciales, se evaluará la bondad de registro y diferenciación espectral mediante el uso de distintas clasificaciones (supervisadas y no supervisadas). Por su influencia directa sobre las observaciones realizadas, se realiza un estudio de la corrección atmosférica; se programan distintos modelos de corrección atmosférica para imágenes multiespectrales y se evalúan los métodos de corrección atmosférica en datos hiperespectrales. Se obtendrá temperatura de la zona de interés utilizando los sensores TM-4, ASTER y OLI, así como un Modelo Digital del Terreno generado por el par estereoscópico capturado por el sensor ASTER. Una vez aplicados estos procedimientos se aplicarán los métodos directos e indirectos, para la localización de zonas probablemente afectadas por la influencia de hidrocarburos y localización directa de hidrocarburos mediante teledetección hiperespectral. Para el método indirecto se utilizan imágenes capturadas por los sensores ETM+ y ASTER. Para el método directo se usan las imágenes capturadas por el sensor Hyperion. ABSTRACT The observation of the Earth is a wonderful tool for studying the different kind of phenomena that occur on its surface. The observation could be done by different scales and by different techniques depending on the information of interest. This Graduate Thesis is intended to expose the territory observation by remote sensing acquiring data systems and the analysis that can be developed to get information of interest. Since Second World War taking aerials photographs of scene was restricted only to a cartographic sense. From these days to nowadays, it have been developed many scientific advances that make capable the interpretation of the surface behavior trough complex systems that are configure by specific geometric and electronic parameters that make possible acquiring time series of the phenomena that manifest on the earth’s surface. Today it is possible to affirm that the exploitation of the electromagnetic spectrum is on a maxim value. In the past, analysis of the electromagnetic spectrum was carry in a narrow part of it, today it is possible to study entire. This implicates the development of new algorithms, process and techniques for the extraction of information about the interaction of matter with electromagnetic radiation. The information that has been acquired by remote sensing sensors is going to be a helpful tool for the exploration of hydrocarbon through direct and vicarious methods. The techniques applied in remote sensing, especially in geophysical campaigns, are employed to minimize costs and maximize results of ground-based geologic investigations. Forecasting of anomalies in the region of interest depends directly on the expertise data analyst who designs, computes and evaluates variations in the electromagnetic energy reflected or emanated from the earth’s surface. For an optimal prediction a review of the capture system take place; assess of the goodness in data acquisition and spectral separability, is carried out by mean of supervised and unsupervised classifications. Due to the direct influence of the atmosphere in the register data, a study of the minimization of its influence has been done; a script has been programed for the atmospheric correction in multispectral data; also, a review of hyperspectral atmospheric correction is conducted. Temperature of the region of interest is computed using the images captured by TM-4, ASTER and OLI, in addition to a Digital Terrain Model generated by a pair of stereo images taken by ASTER sensor. Once these procedures have finished, direct and vicarious methods are applied in order to find altered zones influenced by hydrocarbons, as well as pinpoint directly hydrocarbon presence by mean of hyperspectral remote sensing. For this purpose ETM+ and ASTER sensors are used to apply the vicarious method and Hyperion images are used to apply the direct method.

Relative Concerns on Visible Consumption: A Source of Economic Distortions

Do relative concerns on visible consumption give rise to economic distortions? We re-examine the question posited by Arrow and Dasgupta (2009) building upon their general framework but recognizing that relative concerns can only apply to visible goods (e.g., cars, clothing, jewelry) and that households consume both visible and non-visible goods. Contrary to Arrow and Dasgupta (2009), the answer to this question turns to be always affirmative: the competitive equilibrium will always be different than the socially optimal one, since individuals do not take into account the negative externality they exert on others through the consumption of the visible good, while the social planner does. If one invokes separability assumptions, then the steady state competitive equilibrium consumption of non-visible goods will be strictly lower than the socially optimal one.

Dynamic, convex, and robust optimization with Bayesian learning for response-guided dosing

Thesis (Ph.D.)--University of Washington, 2016-08

Robustness of quantum gates in the presence of noise

We define several quantitative measures of the robustness of a quantum gate against noise. Exact analytic expressions for the robustness against depolarizing noise are obtained for all bipartite unitary quantum gates, and it is found that the controlled-NOT gate is the most robust two-qubit quantum gate, in the sense that it is the quantum gate which can tolerate the most depolarizing noise and still generate entanglement. Our results enable us to place several analytic upper bounds on the value of the threshold for quantum computation, with the best bound in the most pessimistic error model being p(th)less than or equal to0.5.

Entanglement induced by spontaneous emission in spatially extended two-atom systems

The role of the collective antisymmetric state in entanglement creation by spontaneous emission in a system of two non-overlapping two-level atoms has been investigated. Populations of the collective atomic states and the Wootters entanglement measure (concurrence) for two sets of initial atomic conditions are calculated and illustrated graphically. Calculations include the dipole-dipole interaction and a spatial separation between the atoms that the antisymmetric state of the system is included throughout even for small interatomic separations. It is shown that spontaneous emission can lead to a transient entanglement between the atoms even if the atoms were prepared initially in an unentangled state. It is found that the ability of spontaneous emission to create transient entanglement relies on the absence of population in the collective symmetric state of the system. For the initial state of only one atom excited, entanglement builds up rapidly in time and reaches a maximum for parameter values corresponding roughly to zero population in the symmetric state. On the other hand, for the initial condition of both atoms excited, the atoms remain unentangled until the symmetric state is depopulated. A simple physical interpretation of these results is given in terms of the diagonal states of the density matrix of the system. We also study entanglement creation in a system of two non-identical atoms of different transition frequencies. It is found that the entanglement between the atoms can be enhanced compared to that for identical atoms, and can decay with two different time scales resulting from the coherent transfer of the population from the symmetric to the antisymmetric state. In addition, it was found that a decaying initial entanglement between the atoms can display a revival behaviour.

Entanglement and spin squeezing in the two-atom Dicke model

We analyse the relation between the entanglement and spin-squeezing parameter in the two-atom Dicke model and identify the source of the discrepancy recently reported by Banerjee (2001 Preprint quant-ph/0110032) and Zhou et al (2002 J. Opt. B. Quantum Semiclass. Opt. 4 425), namely that one can observe entanglement without spin squeezing. Our calculations demonstrate that there are two criteria for entanglement, one associated with the two-photon coherences that create two-photon entangled states, and the other associated with populations of the collective states. We find that the spin-squeezing parameter correctly predicts entanglement in the two-atom Dicke system only if it is associated with two-photon entangled states, but fails to predict entanglement when it is associated with the entangled symmetric state. This explicitly identifies the source of the discrepancy and explains why the system can be entangled without spin squeezing. We illustrate these findings with three examples of the interaction of the system with thermal, classical squeezed vacuum, and quantum squeezed vacuum fields.

Spin squeezing as a measure of entanglement in a two-qubit system

We show that the two definitions of spin squeezing extensively used in the literature [M. Kitagawa and M. Ueda, Phys. Rev. A 47, 5138 (1993) and D.J. Wineland , Phys. Rev. A 50, 67 (1994)] give different predictions of entanglement in the two-atom Dicke system. We analyze differences between the definitions and show that the spin squeezing parameter of Kitagawa and Ueda is a better measure of entanglement than the commonly used spectroscopic spin squeezing parameter. We illustrate this relation by examining different examples of a driven two-atom Dicke system in which spin squeezing and entanglement arise dynamically. We give an explanation of the source of the difference using the negativity criterion for entanglement.

Multipartite entanglement and quantum state exchange

We investigate multipartite entanglement in relation to the process of quantum state exchange. In particular, we consider such entanglement for a certain pure state involving two groups of N trapped atoms. The state, which can be produced via quantum state exchange, is analogous to the steady-state intracavity state of the subthreshold optical nondegenerate parametric amplifier. We show that, first, it possesses some 2N-way entanglement. Second, we place a lower bound on the amount of such entanglement in the state using a measure called the entanglement of minimum bipartite entropy.

Utility functions on locally connected spaces

We investigate the role of local connectedness in utility theory and prove that any continuous total preorder on a locally connected separable space is continuously representable. This is a new simple criterion for the representability of continuous preferences, and is not a consequence of the standard theorems in utility theory that use conditions such as connectedness and separability, second countability, or path-connectedness. Finally we give applications to problems involving the existence of value functions in population ethics and to the problem of proving the existence of continuous utility functions in general equilibrium models with land as one of the commodities. (C) 2003 Elsevier B.V. All rights reserved.

Experimental characterization of continuous-variable entanglement

We present an experimental analysis of quadrature entanglement produced from a pair of amplitude squeezed beams. The correlation matrix of the state is characterized within a set of reasonable assumptions, and the strength of the entanglement is gauged using measures of the degree of inseparability and the degree of Einstein-Podolsky-Rosen (EPR) paradox. We introduce controlled decoherence in the form of optical loss to the entangled state, and demonstrate qualitative differences in the response of the degrees of inseparability and EPR paradox to this loss. The entanglement is represented on a photon number diagram that provides an intuitive and physically relevant description of the state. We calculate efficacy contours for several quantum information protocols on this diagram, and use them to predict the effectiveness of our entanglement in those protocols.

Stationary two-atom entanglement induced by nonclassical two-photon correlations

A system of two two-level atoms interacting with a squeezed vacuum field can exhibit stationary entanglement associated with nonclassical two-photon correlations characteristic of the squeezed vacuum field. The amount of entanglement present in the system is quantified by the well known measure of entanglement called concurrence. We find analytical formulae describing the concurrence for two identical and nonidentical atoms and show that it is possible to obtain a large degree of steady-state entanglement in the system. Necessary conditions for the entanglement are nonclassical two-photon correlations and nonzero collective decay. It is shown that nonidentical atoms are a better source of stationary entanglement than identical atoms. We discuss the optimal physical conditions for creating entanglement in the system; in particular, it is shown that there is an optimal and rather small value of the mean photon number required for creating entanglement.

Entangling two atoms via spontaneous emission

We discuss the creation of entanglement between two two-level atoms in the dissipative process of spontaneous emission. It is shown that spontaneous emission can lead to a transient entanglement between the atoms even if the atoms were prepared initially in an unentangled state. The amount of entanglement created in the system is quantified by using two different measures: concurrence and negativity. We find analytical formulae for the evolution of concurrence and negativity in the system. We also find the analytical relation between the two measures of entanglement. The system consists of two two-level atoms which are separated by an arbitrary distance r(12) and interact with each other via the dipole-dipole interaction, and the antisymmetric state of the system is included throughout, even for small interatomic separations, in contrast to the small-sample model. It is shown that for sufficiently large values of the dipole-dipole interaction initially the entanglement exhibits oscillatory behaviour with considerable entanglement in the peaks. For longer times the amount of entanglement is directly related to the population of the slowly decaying antisymmetric state.

Detecting multipartite entanglement

We discuss the problem of determining whether the state of several quantum mechanical subsystems is entangled. As in previous work on two subsystems we introduce a procedure for checking separability that is based on finding state extensions with appropriate properties and may be implemented as a semidefinite program. The main result of this work is to show that there is a series of tests of this kind such that if a multiparty state is entangled this will eventually be detected by one of the tests. The procedure also provides a means of constructing entanglement witnesses that could in principle be measured in order to demonstrate that the state is entangled.

Using weak nonlinearity under decoherence for macroscopic entanglement generation and quantum computation

Recently, there have been several suggestions that weak Kerr nonlinearity can be used for generation of macroscopic superpositions and entanglement and for linear optics quantum computation. However, it is not immediately clear that this approach can overcome decoherence effects. Our numerical study shows that nonlinearity of weak strength could be useful for macroscopic entanglement generation and quantum gate operations in the presence of decoherence. We suggest specific values for real experiments based on our analysis. Our discussion shows that the generation of macroscopic entanglement using this approach is within the reach of current technology.