State estimation using model order reduction for unstable systems

The problem of state estimation occurs in many applications of fluid flow. For example, to produce a reliable weather forecast it is essential to find the best possible estimate of the true state of the atmosphere. To find this best estimate a nonlinear least squares problem has to be solved subject to dynamical system constraints. Usually this is solved iteratively by an approximate Gauss–Newton method where the underlying discrete linear system is in general unstable. In this paper we propose a new method for deriving low order approximations to the problem based on a recently developed model reduction method for unstable systems. To illustrate the theoretical results, numerical experiments are performed using a two-dimensional Eady model – a simple model of baroclinic instability, which is the dominant mechanism for the growth of storms at mid-latitudes. It is a suitable test model to show the benefit that may be obtained by using model reduction techniques to approximate unstable systems within the state estimation problem.

Optimal growth in a two-sector model without discounting: a geometric investigation

We characterize optimal policy in a two-sector growth model with xed coeÆcients and with no discounting. The model is a specialization to a single type of machine of a general vintage capital model originally formulated by Robinson, Solow and Srinivasan, and its simplicity is not mirrored in its rich dynamics, and which seem to have been missed in earlier work. Our results are obtained by viewing the model as a specific instance of the general theory of resource allocation as initiated originally by Ramsey and von Neumann and brought to completion by McKenzie. In addition to the more recent literature on chaotic dynamics, we relate our results to the older literature on optimal growth with one state variable: speci cally, to the one-sector setting of Ramsey, Cass and Koopmans, as well as to the two-sector setting of Srinivasan and Uzawa. The analysis is purely geometric, and from a methodological point of view, our work can be seen as an argument, at least in part, for the rehabilitation of geometric methods as an engine of analysis.

Charge-orbital ordering and phase separation in the two-orbital model for manganites: Roles of Jahn-Teller phononic and Coulombic interactions

The main properties of realistic models for manganites are studied using analytic mean-field approximations and computational numerical methods, focusing on the two-orbital model with electrons interacting through Jahn-Teller (JT) phonons and/or Coulombic repulsions. Analyzing the model including both interactions by the combination of the mean-field approximation and the exact diagonalization method, it is argued that the spin-charge-orbital structure in the insulating phase of the purely JT-phononic model with a large Hund couphng J(H) is not qualitatively changed by the inclusion of the Coulomb interactions. As an important application of the present mean-held approximation, the CE-type antiferromagnetic state, the charge-stacked structure along the z axis, and (3x(2) - r(2))/(3y(2) - r(2))-type orbital ordering are successfully reproduced based on the JT-phononic model with large JH for the half-doped manganite, in agreement with recent Monte Carlo simulation results. Topological arguments and the relevance of the Heisenberg exchange among localized t(2g) spins explains why the inclusion of the nearest-neighbor Coulomb interaction does not destroy the charge stacking structure. It is also verified that the phase-separation tendency is observed both in purely JT-phononic (large JH) and purely Coulombic models in the vicinity of the hole undoped region, as long as realistic hopping matrices are used. This highlights the qualitative similarities of both approaches and the relevance of mixed-phase tendencies in the context of manganites. In addition, the rich and complex phase diagram of the two-orbital Coulombic model in one dimension is presented. Our results provide robust evidence that Coulombic and JT-phononic approaches to manganites are not qualitatively different ways to carry out theoretical calculations, but they share a variety of common features.

Biomedical events extraction using the hidden vector state model

Objective: Biomedical events extraction concerns about events describing changes on the state of bio-molecules from literature. Comparing to the protein-protein interactions (PPIs) extraction task which often only involves the extraction of binary relations between two proteins, biomedical events extraction is much harder since it needs to deal with complex events consisting of embedded or hierarchical relations among proteins, events, and their textual triggers. In this paper, we propose an information extraction system based on the hidden vector state (HVS) model, called HVS-BioEvent, for biomedical events extraction, and investigate its capability in extracting complex events. Methods and material: HVS has been previously employed for extracting PPIs. In HVS-BioEvent, we propose an automated way to generate abstract annotations for HVS training and further propose novel machine learning approaches for event trigger words identification, and for biomedical events extraction from the HVS parse results. Results: Our proposed system achieves an F-score of 49.57% on the corpus used in the BioNLP'09 shared task, which is only 2.38% lower than the best performing system by UTurku in the BioNLP'09 shared task. Nevertheless, HVS-BioEvent outperforms UTurku's system on complex events extraction with 36.57% vs. 30.52% being achieved for extracting regulation events, and 40.61% vs. 38.99% for negative regulation events. Conclusions: The results suggest that the HVS model with the hierarchical hidden state structure is indeed more suitable for complex event extraction since it could naturally model embedded structural context in sentences.

Discriminative training of the hidden vector state model for semantic parsing

In this paper, we discuss how discriminative training can be applied to the hidden vector state (HVS) model in different task domains. The HVS model is a discrete hidden Markov model (HMM) in which each HMM state represents the state of a push-down automaton with a finite stack size. In previous applications, maximum-likelihood estimation (MLE) is used to derive the parameters of the HVS model. However, MLE makes a number of assumptions and unfortunately some of these assumptions do not hold. Discriminative training, without making such assumptions, can improve the performance of the HVS model by discriminating the correct hypothesis from the competing hypotheses. Experiments have been conducted in two domains: the travel domain for the semantic parsing task using the DARPA Communicator data and the Air Travel Information Services (ATIS) data and the bioinformatics domain for the information extraction task using the GENIA corpus. The results demonstrate modest improvements of the performance of the HVS model using discriminative training. In the travel domain, discriminative training of the HVS model gives a relative error reduction rate of 31 percent in F-measure when compared with MLE on the DARPA Communicator data and 9 percent on the ATIS data. In the bioinformatics domain, a relative error reduction rate of 4 percent in F-measure is achieved on the GENIA corpus.

Theoretical study of one- and two-photon absorption spectra of azoaromatic compounds

In this study, the one- and two-photon absorption spectra of seven azoaromatic compounds (five pseudostilbenes-type and two aminoazobenzenes) were theoretically investigated using the density functional theory combined with the response functions formalism. The equilibrium molecular structure of each compound was obtained at three different levels of theory: Hartree-Fock, density functional theory (DFT), and Moller-Plesset 2. The effect of solvent on the equilibrium structure and the electronic transitions of the compounds were investigated using the polarizable continuum model. For the one-photon absorption, the allowed pi ->pi(*) transition energy showed to be dependent on the molecular structures and the effect of solvent, while the n ->pi(*) and pi ->pi(*)(n) transition energies exhibited only a slight dependence. An inversion between the bands corresponding to the pi ->pi(*) and n ->pi(*) states due to the effect of solvent was observed for the pseudostilbene-type compounds. To characterize the allowed two-photon absorption transitions for azoaromatic compounds, the response functions formalism combined with DFT using the hybrid B3LYP and PBE0 functionals and the long-range corrected CAM-B3LYP functional was employed. The theoretical results support the previous findings based on the three-state model. The model takes into account the ground and two electronic excited states and has already been used to describe and interpret the two-photon absorption spectrum of azoaromatic compounds. The highest energy two-photon allowed transition for the pseudostilbene-type compounds shows to be more effectively affected (similar to 20%) by the torsion of the molecular structure than the lowest allowed transition (similar to 10%). In order to elucidate the effect of the solvent on the two-photon absorption spectra, the lowest allowed two-photon transition (dipolar transition) for each compound was analyzed using a two-state approximation and the polarizable continuum model. The results obtained reveal that the effect of solvent increases drastically the two-photon cross-section of the dipolar transition of the pseudostilbene-type compounds. In general, the features of both one- and two-photon absorption spectra of the azoaromatic compounds are well reproduced by the theoretical calculations.

Nine classes of integrable boundary conditions for the eight-state supersymmetric fermion model

Nine classes of integrable boundary conditions for the eight-state supersymmetric model of strongly correlated fermions are presented. The boundary systems are solved by using the coordinate Bethe ansatz method and the Bethe ansatz equations for all nine cases are given.

Use of Two-Part Regression Calibration Model to Correct for Measurement Error in Episodically Consumed Foods in a Single-Replicate Study Design: EPIC Case Study.

In epidemiologic studies, measurement error in dietary variables often attenuates association between dietary intake and disease occurrence. To adjust for the attenuation caused by error in dietary intake, regression calibration is commonly used. To apply regression calibration, unbiased reference measurements are required. Short-term reference measurements for foods that are not consumed daily contain excess zeroes that pose challenges in the calibration model. We adapted two-part regression calibration model, initially developed for multiple replicates of reference measurements per individual to a single-replicate setting. We showed how to handle excess zero reference measurements by two-step modeling approach, how to explore heteroscedasticity in the consumed amount with variance-mean graph, how to explore nonlinearity with the generalized additive modeling (GAM) and the empirical logit approaches, and how to select covariates in the calibration model. The performance of two-part calibration model was compared with the one-part counterpart. We used vegetable intake and mortality data from European Prospective Investigation on Cancer and Nutrition (EPIC) study. In the EPIC, reference measurements were taken with 24-hour recalls. For each of the three vegetable subgroups assessed separately, correcting for error with an appropriately specified two-part calibration model resulted in about three fold increase in the strength of association with all-cause mortality, as measured by the log hazard ratio. Further found is that the standard way of including covariates in the calibration model can lead to over fitting the two-part calibration model. Moreover, the extent of adjusting for error is influenced by the number and forms of covariates in the calibration model. For episodically consumed foods, we advise researchers to pay special attention to response distribution, nonlinearity, and covariate inclusion in specifying the calibration model.

Lattice-gas model of orientable molecules: Application to liquid crystals

We have analyzed a two-dimensional lattice-gas model of cylindrical molecules which can exhibit four possible orientations. The Hamiltonian of the model contains positional and orientational energy interaction terms. The ground state of the model has been investigated on the basis of Karl¿s theorem. Monte Carlo simulation results have confirmed the predicted ground state. The model is able to reproduce, with appropriate values of the Hamiltonian parameters, both, a smectic-nematic-like transition and a nematic-isotropic-like transition. We have also analyzed the phase diagram of the system by mean-field techniques and Monte Carlo simulations. Mean-field calculations agree well qualitatively with Monte Carlo results but overestimate transition temperatures.

Lien entre les matrices de transfert de spins et de boucles du modèle de Potts sur le tore

Le lien entre le spectre de la matrice de transfert de la formulation de spins du modèle de Potts critique et celui de la matrice de transfert double-ligne de la formulation de boucles est établi. La relation entre la trace des deux opérateurs est obtenue dans deux représentations de l'algèbre de Temperley-Lieb cyclique, dont la matrice de transfert de boucles est un élément. Le résultat est exprimé en termes des traces modifiées, qui correspondent à des traces effectuées dans le sous-espace de l'espace de représentation des N-liens se transformant selon la m ième représentation irréductible du groupe cyclique. Le mémoire comporte trois chapitres. Dans le premier chapitre, les résultats essentiels concernant les formulations de spins et de boucles du modèle de Potts sont rappelés. Dans le second chapitre, les propriétés de l'algèbre de Temperley-Lieb cyclique et de ses représentations sont étudiées. Enfin, le lien entre les deux traces est construit dans le troisième chapitre. Le résultat final s'apparente à celui obtenu par Richard et Jacobsen en 2007, mais une nouvelle représentation n'ayant pas été étudiée est aussi investiguée.

Essays on numerically efficient inference in nonlinear and non-Gaussian state space models, and commodity market analysis.

The first two articles build procedures to simulate vector of univariate states and estimate parameters in nonlinear and non Gaussian state space models. We propose state space speci fications that offer more flexibility in modeling dynamic relationship with latent variables. Our procedures are extension of the HESSIAN method of McCausland[2012]. Thus, they use approximation of the posterior density of the vector of states that allow to : simulate directly from the state vector posterior distribution, to simulate the states vector in one bloc and jointly with the vector of parameters, and to not allow data augmentation. These properties allow to build posterior simulators with very high relative numerical efficiency. Generic, they open a new path in nonlinear and non Gaussian state space analysis with limited contribution of the modeler. The third article is an essay in commodity market analysis. Private firms coexist with farmers' cooperatives in commodity markets in subsaharan african countries. The private firms have the biggest market share while some theoretical models predict they disappearance once confronted to farmers cooperatives. Elsewhere, some empirical studies and observations link cooperative incidence in a region with interpersonal trust, and thus to farmers trust toward cooperatives. We propose a model that sustain these empirical facts. A model where the cooperative reputation is a leading factor determining the market equilibrium of a price competition between a cooperative and a private firm

Comprehensive evaluation of the WaterGAP3 model across climatic, physiographic, and anthropogenic gradients

In den letzten Jahrzehnten haben sich makroskalige hydrologische Modelle als wichtige Werkzeuge etabliert um den Zustand der globalen erneuerbaren Süßwasserressourcen flächendeckend bewerten können. Sie werden heutzutage eingesetzt um eine große Bandbreite wissenschaftlicher Fragestellungen zu beantworten, insbesondere hinsichtlich der Auswirkungen anthropogener Einflüsse auf das natürliche Abflussregime oder der Auswirkungen des globalen Wandels und Klimawandels auf die Ressource Wasser. Diese Auswirkungen lassen sich durch verschiedenste wasserbezogene Kenngrößen abschätzen, wie z.B. erneuerbare (Grund-)Wasserressourcen, Hochwasserrisiko, Dürren, Wasserstress und Wasserknappheit. Die Weiterentwicklung makroskaliger hydrologischer Modelle wurde insbesondere durch stetig steigende Rechenkapazitäten begünstigt, aber auch durch die zunehmende Verfügbarkeit von Fernerkundungsdaten und abgeleiteten Datenprodukten, die genutzt werden können, um die Modelle anzutreiben und zu verbessern. Wie alle makro- bis globalskaligen Modellierungsansätze unterliegen makroskalige hydrologische Simulationen erheblichen Unsicherheiten, die (i) auf räumliche Eingabedatensätze, wie z.B. meteorologische Größen oder Landoberflächenparameter, und (ii) im Besonderen auf die (oftmals) vereinfachte Abbildung physikalischer Prozesse im Modell zurückzuführen sind. Angesichts dieser Unsicherheiten ist es unabdingbar, die tatsächliche Anwendbarkeit und Prognosefähigkeit der Modelle unter diversen klimatischen und physiographischen Bedingungen zu überprüfen. Bisher wurden die meisten Evaluierungsstudien jedoch lediglich in wenigen, großen Flusseinzugsgebieten durchgeführt oder fokussierten auf kontinentalen Wasserflüssen. Dies steht im Kontrast zu vielen Anwendungsstudien, deren Analysen und Aussagen auf simulierten Zustandsgrößen und Flüssen in deutlich feinerer räumlicher Auflösung (Gridzelle) basieren. Den Kern der Dissertation bildet eine umfangreiche Evaluierung der generellen Anwendbarkeit des globalen hydrologischen Modells WaterGAP3 für die Simulation von monatlichen Abflussregimen und Niedrig- und Hochwasserabflüssen auf Basis von mehr als 2400 Durchflussmessreihen für den Zeitraum 1958-2010. Die betrachteten Flusseinzugsgebiete repräsentieren ein breites Spektrum klimatischer und physiographischer Bedingungen, die Einzugsgebietsgröße reicht von 3000 bis zu mehreren Millionen Quadratkilometern. Die Modellevaluierung hat dabei zwei Zielsetzungen: Erstens soll die erzielte Modellgüte als Bezugswert dienen gegen den jegliche weiteren Modellverbesserungen verglichen werden können. Zweitens soll eine Methode zur diagnostischen Modellevaluierung entwickelt und getestet werden, die eindeutige Ansatzpunkte zur Modellverbesserung aufzeigen soll, falls die Modellgüte unzureichend ist. Hierzu werden komplementäre Modellgütemaße mit neun Gebietsparametern verknüpft, welche die klimatischen und physiographischen Bedingungen sowie den Grad anthropogener Beeinflussung in den einzelnen Einzugsgebieten quantifizieren. WaterGAP3 erzielt eine mittlere bis hohe Modellgüte für die Simulation von sowohl monatlichen Abflussregimen als auch Niedrig- und Hochwasserabflüssen, jedoch sind für alle betrachteten Modellgütemaße deutliche räumliche Muster erkennbar. Von den neun betrachteten Gebietseigenschaften weisen insbesondere der Ariditätsgrad und die mittlere Gebietsneigung einen starken Einfluss auf die Modellgüte auf. Das Modell tendiert zur Überschätzung des jährlichen Abflussvolumens mit steigender Aridität. Dieses Verhalten ist charakteristisch für makroskalige hydrologische Modelle und ist auf die unzureichende Abbildung von Prozessen der Abflussbildung und –konzentration in wasserlimitierten Gebieten zurückzuführen. In steilen Einzugsgebieten wird eine geringe Modellgüte hinsichtlich der Abbildung von monatlicher Abflussvariabilität und zeitlicher Dynamik festgestellt, die sich auch in der Güte der Niedrig- und Hochwassersimulation widerspiegelt. Diese Beobachtung weist auf notwendige Modellverbesserungen in Bezug auf (i) die Aufteilung des Gesamtabflusses in schnelle und verzögerte Abflusskomponente und (ii) die Berechnung der Fließgeschwindigkeit im Gerinne hin. Die im Rahmen der Dissertation entwickelte Methode zur diagnostischen Modellevaluierung durch Verknüpfung von komplementären Modellgütemaßen und Einzugsgebietseigenschaften wurde exemplarisch am Beispiel des WaterGAP3 Modells erprobt. Die Methode hat sich als effizientes Werkzeug erwiesen, um räumliche Muster in der Modellgüte zu erklären und Defizite in der Modellstruktur zu identifizieren. Die entwickelte Methode ist generell für jedes hydrologische Modell anwendbar. Sie ist jedoch insbesondere für makroskalige Modelle und multi-basin Studien relevant, da sie das Fehlen von feldspezifischen Kenntnissen und gezielten Messkampagnen, auf die üblicherweise in der Einzugsgebietsmodellierung zurückgegriffen wird, teilweise ausgleichen kann.

An integrative model of amino acid metabolism in the liver of the lactating dairy cow

The objective of this work was to construct a dynamic model of hepatic amino acid metabolism in the lactating dairy cow that could be parameterized using net flow data from in vivo experiments. The model considers 22 amino acids, ammonia, urea, and 13 energetic metabolites, and was parameterized using a steady-state balance model and two in vivo, net flow experiments conducted with mid-lactation dairy cows. Extracellular flows were derived directly from the observed data. An optimization routine was used to derive nine intracellular flows. The resulting dynamic model was found to be stable across a range of inputs suggesting that it can be perturbed and applied to other physiological states. Although nitrogen was generally in balance, leucine was in slight deficit compared to predicted needs for export protein synthesis, suggesting that an alternative source of leucine (e.g. peptides) was utilized. Simulations of varying glucagon concentrations indicated that an additional 5 mol/d of glucose could be synthesized at the reference substrate concentrations and blood flows. The increased glucose production was supported by increased removal from blood of lactate, glutamate, aspartate, alanine, asparagine, and glutamine. As glucose Output increased, ketone body and acetate release increased while CO2 release declined. The pattern of amino acids appearing in hepatic vein blood was affected by changes in amino acid concentration in portal vein blood, portal blood flow rate and glucagon concentration, with methionine and phenylalanine being the most affected of essential amino acids. Experimental evidence is insufficient to determine whether essential amino acids are affected by varying gluconeogenic demands. (C) 2004 Published by Elsevier Ltd.

Pseudolikelihood equations for Potts MRF model parameter estimation on higher order neighborhood systems

This letter presents pseudolikelihood equations for the estimation of the Potts Markov random field model parameter on higher order neighborhood systems. The derived equation for second-order systems is a significantly reduced version of a recent result found in the literature (from 67 to 22 terms). Also, with the proposed method, a completely original equation for Potts model parameter estimation in third-order systems was obtained. These equations allow the modeling of less restrictive contextual systems for a large number of applications in a computationally feasible way. Experiments with both simulated and real remote sensing images provided good results.

On the stability of the extracellular hemoglobin of Glossoscolex paulistus, in two iron oxidation states, in the presence of urea

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