Hilbert Space Valued Signal Plus Noise Models: Analysis of Structural Breaks under High Dimensionality and Temporal Dependence

This thesis is concerned with change point analysis for time series, i.e. with detection of structural breaks in time-ordered, random data. This long-standing research field regained popularity over the last few years and is still undergoing, as statistical analysis in general, a transformation to high-dimensional problems. We focus on the fundamental »change in the mean« problem and provide extensions of the classical non-parametric Darling-Erdős-type cumulative sum (CUSUM) testing and estimation theory within highdimensional Hilbert space settings. In the first part we contribute to (long run) principal component based testing methods for Hilbert space valued time series under a rather broad (abrupt, epidemic, gradual, multiple) change setting and under dependence. For the dependence structure we consider either traditional m-dependence assumptions or more recently developed m-approximability conditions which cover, e.g., MA, AR and ARCH models. We derive Gumbel and Brownian bridge type approximations of the distribution of the test statistic under the null hypothesis of no change and consistency conditions under the alternative. A new formulation of the test statistic using projections on subspaces allows us to simplify the standard proof techniques and to weaken common assumptions on the covariance structure. Furthermore, we propose to adjust the principal components by an implicit estimation of a (possible) change direction. This approach adds flexibility to projection based methods, weakens typical technical conditions and provides better consistency properties under the alternative. In the second part we contribute to estimation methods for common changes in the means of panels of Hilbert space valued time series. We analyze weighted CUSUM estimates within a recently proposed »high-dimensional low sample size (HDLSS)« framework, where the sample size is fixed but the number of panels increases. We derive sharp conditions on »pointwise asymptotic accuracy« or »uniform asymptotic accuracy« of those estimates in terms of the weighting function. Particularly, we prove that a covariance-based correction of Darling-Erdős-type CUSUM estimates is required to guarantee uniform asymptotic accuracy under moderate dependence conditions within panels and that these conditions are fulfilled, e.g., by any MA(1) time series. As a counterexample we show that for AR(1) time series, close to the non-stationary case, the dependence is too strong and uniform asymptotic accuracy cannot be ensured. Finally, we conduct simulations to demonstrate that our results are practically applicable and that our methodological suggestions are advantageous.

Validez de contenido de un cuestionario para evaluar las etapas del cambio de comportamiento en estilos de vida en escolares, Colombia: Estudio FUPRECOL

Estudio de validación en escolares pertenecientes a instituciones educativas oficiales de la ciudad de Bogotá, Colombia. Se diseñó y aplicó el CCC-FUPRECOL que indagó por las etapas de cambio para la actividad física/ejercicio, consumo de frutas, verduras, drogas, tabaco e ingesta de bebidas alcohólicas, de manera auto-diligenciada por formulario estructurado.

Influence of data structure on the estimation of the additive genetic direct and maternal covariance for early growth traits in Nellore cattle

The objective of the present study was to investigate the effect of data structure on estimated genetic parameters and predicted breeding values of direct and maternal genetic effects for weaning weight (WW) and weight gain from birth to weaning (BWG), including or not the genetic covariance between direct and maternal effects. Records of 97,490 Nellore animals born between 1993 and 2006, from the Jacarezinho cattle raising farm, were used. Two different data sets were analyzed: DI_all, which included all available progenies of dams without their own performance; DII_all, which included DI_all + 20% of recorded progenies with maternal phenotypes. Two subsets were obtained from each data set (DI_all and DII_all): DI_1 and DII_1, which included only dams with three or fewer progenies; DI_5 and DII_5, which included only dams with five or more progenies. (Co)variance components and heritabilities were estimated by Bayesian inference through Gibbs sampling using univariate animal models. In general, for the population and traits studied, the proportion of dams with known phenotypic information and the number of progenies per dam influenced direct and maternal heritabilities, as well as the contribution of maternal permanent environmental variance to phenotypic variance. Only small differences were observed in the genetic and environmental parameters when the genetic covariance between direct and maternal effects was set to zero in the data sets studied. Thus, the inclusion or not of the genetic covariance between direct and maternal effects had little effect on the ranking of animals according to their breeding values for WW and BWG. Accurate estimation of genetic correlations between direct and maternal genetic effects depends on the data structure. Thus, this covariance should be set to zero in Nellore data sets in which the proportion of dams with phenotypic information is low, the number of progenies per dam is small, and pedigree relationships are poorly known. (c) 2012 Elsevier B.V. All rights reserved.

Estimates of covariance functions for growth of Nelore cattle applying a parametric correlation structure to model within-animal correlations

A total of 20,065 weights recorded on 3016 Nelore animals were used to estimate covariance functions for growth from birth to 630 days of age, assuming a parametric correlation structure to model within-animal correlations. The model of analysis included fixed effects of contemporary groups and age of dam as quadratic covariable. Mean trends were taken into account by a cubic regression on orthogonal polynomials of animal age. Genetic effects of the animal and its dam and maternal permanent environmental effects were modelled by random regressions on Legendre polynomials of age at recording. Changes in direct permanent environmental effect variances were modelled by a polynomial variance function, together with a parametric correlation function to account for correlations between ages. Stationary and nonstationary models were used to model within-animal correlations between different ages. Residual variances were considered homogeneous or heterogeneous, with changes modelled by a step or polynomial function of age at recording. Based on Bayesian information criterion, a model with a cubic variance function combined with a nonstationary correlation function for permanent environmental effects, with 49 parameters to be estimated, fitted best. Modelling within-animal correlations through a parametric correlation structure can describe the variation pattern adequately. Moreover, the number of parameters to be estimated can be decreased substantially compared to a model fitting random regression on Legendre polynomial of age. © 2004 Elsevier B.V. All rights reserved.

Modeling strategies in longitudinal data analysis: Covariate, variance function and correlation structure selection

A modeling paradigm is proposed for covariate, variance and working correlation structure selection for longitudinal data analysis. Appropriate selection of covariates is pertinent to correct variance modeling and selecting the appropriate covariates and variance function is vital to correlation structure selection. This leads to a stepwise model selection procedure that deploys a combination of different model selection criteria. Although these criteria find a common theoretical root based on approximating the Kullback-Leibler distance, they are designed to address different aspects of model selection and have different merits and limitations. For example, the extended quasi-likelihood information criterion (EQIC) with a covariance penalty performs well for covariate selection even when the working variance function is misspecified, but EQIC contains little information on correlation structures. The proposed model selection strategies are outlined and a Monte Carlo assessment of their finite sample properties is reported. Two longitudinal studies are used for illustration.

Working-correlation-structure identification in generalized estimating equations

Selecting an appropriate working correlation structure is pertinent to clustered data analysis using generalized estimating equations (GEE) because an inappropriate choice will lead to inefficient parameter estimation. We investigate the well-known criterion of QIC for selecting a working correlation Structure. and have found that performance of the QIC is deteriorated by a term that is theoretically independent of the correlation structures but has to be estimated with an error. This leads LIS to propose a correlation information criterion (CIC) that substantially improves the QIC performance. Extensive simulation studies indicate that the CIC has remarkable improvement in selecting the correct correlation structures. We also illustrate our findings using a data set from the Madras Longitudinal Schizophrenia Study.

Working correlation structure misspecification, estimation and covariate design: Implications for generalised estimating equations performance

The method of generalised estimating equations for regression modelling of clustered outcomes allows for specification of a working matrix that is intended to approximate the true correlation matrix of the observations. We investigate the asymptotic relative efficiency of the generalised estimating equation for the mean parameters when the correlation parameters are estimated by various methods. The asymptotic relative efficiency depends on three-features of the analysis, namely (i) the discrepancy between the working correlation structure and the unobservable true correlation structure, (ii) the method by which the correlation parameters are estimated and (iii) the 'design', by which we refer to both the structures of the predictor matrices within clusters and distribution of cluster sizes. Analytical and numerical studies of realistic data-analysis scenarios show that choice of working covariance model has a substantial impact on regression estimator efficiency. Protection against avoidable loss of efficiency associated with covariance misspecification is obtained when a 'Gaussian estimation' pseudolikelihood procedure is used with an AR(1) structure.

Stationary processes for object detection and non-rigid structure-from-motion

Stationary processes are random variables whose value is a signal and whose distribution is invariant to translation in the domain of the signal. They are intimately connected to convolution, and therefore to the Fourier transform, since the covariance matrix of a stationary process is a Toeplitz matrix, and Toeplitz matrices are the expression of convolution as a linear operator. This thesis utilises this connection in the study of i) efficient training algorithms for object detection and ii) trajectory-based non-rigid structure-from-motion.

Disjunct eddy covariance measurements of volatile organic compound fluxes using proton transfer reaction mass spectrometry

Volatile organic compounds (VOCs) are emitted into the atmosphere from natural and anthropogenic sources, vegetation being the dominant source on a global scale. Some of these reactive compounds are deemed major contributors or inhibitors to aerosol particle formation and growth, thus making VOC measurements essential for current climate change research. This thesis discusses ecosystem scale VOC fluxes measured above a boreal Scots pine dominated forest in southern Finland. The flux measurements were performed using the micrometeorological disjunct eddy covariance (DEC) method combined with proton transfer reaction mass spectrometry (PTR-MS), which is an online technique for measuring VOC concentrations. The measurement, calibration, and calculation procedures developed in this work proved to be well suited to long-term VOC concentration and flux measurements with PTR-MS. A new averaging approach based on running averaged covariance functions improved the determination of the lag time between wind and concentration measurements, which is a common challenge in DEC when measuring fluxes near the detection limit. The ecosystem scale emissions of methanol, acetaldehyde, and acetone were substantial. These three oxygenated VOCs made up about half of the total emissions, with the rest comprised of monoterpenes. Contrary to the traditional assumption that monoterpene emissions from Scots pine originate mainly as evaporation from specialized storage pools, the DEC measurements indicated a significant contribution from de novo biosynthesis to the ecosystem scale monoterpene emissions. This thesis offers practical guidelines for long-term DEC measurements with PTR-MS. In particular, the new averaging approach to the lag time determination seems useful in the automation of DEC flux calculations. Seasonal variation in the monoterpene biosynthesis and the detailed structure of a revised hybrid algorithm, describing both de novo and pool emissions, should be determined in further studies to improve biological realism in the modelling of monoterpene emissions from Scots pine forests. The increasing number of DEC measurements of oxygenated VOCs will probably enable better estimates of the role of these compounds in plant physiology and tropospheric chemistry. Keywords: disjunct eddy covariance, lag time determination, long-term flux measurements, proton transfer reaction mass spectrometry, Scots pine forests, volatile organic compounds

Covariance profiles: a signature representation for object sets

We consider the problem of extracting a signature representation of similar entities employing covariance descriptors. Covariance descriptors can efficiently represent objects and are robust to scale and pose changes. We posit that covariance descriptors corresponding to similar objects share a common geometrical structure which can be extracted through joint diagonalization. We term this diagonalizing matrix as the Covariance Profile (CP). CP can be used to measure the distance of a novel object to an object set through the diagonality measure. We demonstrate how CP can be employed on images as well as for videos, for applications such as face recognition and object-track clustering.

Simulating dynamic covariance structures for testing the adaptive behavior of variable selection algorithms (Invited Paper)

Variable selection for regression is a classical statistical problem, motivated by concerns that too large a number of covariates may bring about overfitting and unnecessarily high measurement costs. Novel difficulties arise in streaming contexts, where the correlation structure of the process may be drifting, in which case it must be constantly tracked so that selections may be revised accordingly. A particularly interesting phenomenon is that non-selected covariates become missing variables, inducing bias on subsequent decisions. This raises an intricate exploration-exploitation tradeoff, whose dependence on the covariance tracking algorithm and the choice of variable selection scheme is too complex to be dealt with analytically. We hence capitalise on the strength of simulations to explore this problem, taking the opportunity to tackle the difficult task of simulating dynamic correlation structures. © 2008 IEEE.

Streaming covariance selection with applications to adaptive querying in sensor networks

Sensor networks can be naturally represented as graphical models, where the edge set encodes the presence of sparsity in the correlation structure between sensors. Such graphical representations can be valuable for information mining purposes as well as for optimizing bandwidth and battery usage with minimal loss of estimation accuracy. We use a computationally efficient technique for estimating sparse graphical models which fits a sparse linear regression locally at each node of the graph via the Lasso estimator. Using a recently suggested online, temporally adaptive implementation of the Lasso, we propose an algorithm for streaming graphical model selection over sensor networks. With battery consumption minimization applications in mind, we use this algorithm as the basis of an adaptive querying scheme. We discuss implementation issues in the context of environmental monitoring using sensor networks, where the objective is short-term forecasting of local wind direction. The algorithm is tested against real UK weather data and conclusions are drawn about certain tradeoffs inherent in decentralized sensor networks data analysis. © 2010 The Author. Published by Oxford University Press on behalf of The British Computer Society. All rights reserved.

The factor structure of the revised Illness Perceptions Questionnaire in a population of oesophageal cancer survivors

Objectives
To determine whether the proposed 7-factor structure of the Illness Perception Questionnaire-Revised (Timeline Acute/Chronic, Timeline Cyclical, Consequences, Personal Control, Treatment Control, Illness Coherence and Emotional Representations) is appropriate among a population of oesophageal cancer survivors.
Methods
Everyone registered with the Oesophageal Patients’ Association in the UK (n=2185) was mailed a questionnaire booklet which included the Illness Perception Questionnaire-Revised. Responses from 587 oesophageal cancer survivors (27%) were subjected to a confirmatory factor analysis.
Results
The proposed 7 factor structure provided a reasonable fit of the data. Modification indices suggested that a significantly better fit could be provided if one of the items on the Timeline Acute/Chronic factor loaded on the Treatment Control factor and an error covariance was added between 2 other items on the Timeline Acute/Chronic factor.
Conclusions
The model fit for the 7 factor structure proposed by Moss-Morris et al. (2002) was found to be adequate in our study. However, the structure of the timeline acute/chronic factor needs to be considered, particularly when the IPQ-R is to be used among older people with a potentially life-threatening illness or those receiving palliative care.

Are power laws that estimate fractal dimension a good descriptor of soil structure and its link to soil biological properties?

The study of interrelationships between soil structure and its functional properties is complicated by the fact that the quantitative description of soil structure is challenging. Soil scientists have tackled this challenge by taking advantage of approaches such as fractal geometry, which describes soil architectural complexity through a scaling exponent (D) relating mass and numbers of particles/aggregates to particle/aggregate size. Typically, soil biologists use empirical indices such as mean weight diameters (MWD) and percent of water stable aggregates (WSA), or the entire size distribution, and they have successfully related these indices to key soil features such as C and N dynamics and biological promoters of soil structure. Here, we focused on D, WSA and MWD and we tested whether: D estimated by the exponent of the power law of number-size distributions is a good and consistent correlate of MWD and WSA; D carries information that differs from MWD and WSA; the fraction of variation in D that is uncorrelated with MWD and WSA is related to soil chemical and biological properties that are thought to establish interdependence with soil structure (e.g., organic C, N, arbuscular mycorrhizal fungi). We analysed observational data from a broad scale field study and results from a greenhouse experiment where arbuscular mycorrhizal fungi (AMF) and collembola altered soil structure. We were able to develop empirical models that account for a highly significant and large portion of the correlation observed between WSA and MWD but we did not uncover the mechanisms that underlie this correlation. We conclude that most of the covariance between D and soil biotic (AMF, plant roots) and abiotic (C. N) properties can be accounted for by WSA and MWD. This result implies that the ecological effects of the fragmentation properties described by D and generally discussed under the framework of fractal models can be interpreted under the intuitive perspective of simpler indices and we suggest that the biotic components mostly impacted the largest size fractions, which dominate MWD, WSA and the scaling exponent ruling number-size distributions. (C) 2010 Elsevier Ltd. All rights reserved.

Quantification de la relation séquence-activité de l’ARN par prédiction de structure tridimensionnelle

Dans un premier temps, nous avons modélisé la structure d’une famille d’ARN avec une grammaire de graphes afin d’identifier les séquences qui en font partie. Plusieurs autres méthodes de modélisation ont été développées, telles que des grammaires stochastiques hors-contexte, des modèles de covariance, des profils de structures secondaires et des réseaux de contraintes. Ces méthodes de modélisation se basent sur la structure secondaire classique comparativement à nos grammaires de graphes qui se basent sur les motifs cycliques de nucléotides. Pour exemplifier notre modèle, nous avons utilisé la boucle E du ribosome qui contient le motif Sarcin-Ricin qui a été largement étudié depuis sa découverte par cristallographie aux rayons X au début des années 90. Nous avons construit une grammaire de graphes pour la structure du motif Sarcin-Ricin et avons dérivé toutes les séquences qui peuvent s’y replier. La pertinence biologique de ces séquences a été confirmée par une comparaison des séquences d’un alignement de plus de 800 séquences ribosomiques bactériennes. Cette comparaison a soulevée des alignements alternatifs pour quelques unes des séquences que nous avons supportés par des prédictions de structures secondaires et tertiaires. Les motifs cycliques de nucléotides ont été observés par les membres de notre laboratoire dans l'ARN dont la structure tertiaire a été résolue expérimentalement. Une étude des séquences et des structures tertiaires de chaque cycle composant la structure du Sarcin-Ricin a révélé que l'espace des séquences dépend grandement des interactions entre tous les nucléotides à proximité dans l’espace tridimensionnel, c’est-à-dire pas uniquement entre deux paires de bases adjacentes. Le nombre de séquences générées par la grammaire de graphes est plus petit que ceux des méthodes basées sur la structure secondaire classique. Cela suggère l’importance du contexte pour la relation entre la séquence et la structure, d’où l’utilisation d’une grammaire de graphes contextuelle plus expressive que les grammaires hors-contexte. Les grammaires de graphes que nous avons développées ne tiennent compte que de la structure tertiaire et négligent les interactions de groupes chimiques spécifiques avec des éléments extra-moléculaires, comme d’autres macromolécules ou ligands. Dans un deuxième temps et pour tenir compte de ces interactions, nous avons développé un modèle qui tient compte de la position des groupes chimiques à la surface des structures tertiaires. L’hypothèse étant que les groupes chimiques à des positions conservées dans des séquences prédéterminées actives, qui sont déplacés dans des séquences inactives pour une fonction précise, ont de plus grandes chances d’être impliqués dans des interactions avec des facteurs. En poursuivant avec l’exemple de la boucle E, nous avons cherché les groupes de cette boucle qui pourraient être impliqués dans des interactions avec des facteurs d'élongation. Une fois les groupes identifiés, on peut prédire par modélisation tridimensionnelle les séquences qui positionnent correctement ces groupes dans leurs structures tertiaires. Il existe quelques modèles pour adresser ce problème, telles que des descripteurs de molécules, des matrices d’adjacences de nucléotides et ceux basé sur la thermodynamique. Cependant, tous ces modèles utilisent une représentation trop simplifiée de la structure d’ARN, ce qui limite leur applicabilité. Nous avons appliqué notre modèle sur les structures tertiaires d’un ensemble de variants d’une séquence d’une instance du Sarcin-Ricin d’un ribosome bactérien. L’équipe de Wool à l’université de Chicago a déjà étudié cette instance expérimentalement en testant la viabilité de 12 variants. Ils ont déterminé 4 variants viables et 8 létaux. Nous avons utilisé cet ensemble de 12 séquences pour l’entraînement de notre modèle et nous avons déterminé un ensemble de propriétés essentielles à leur fonction biologique. Pour chaque variant de l’ensemble d’entraînement nous avons construit des modèles de structures tertiaires. Nous avons ensuite mesuré les charges partielles des atomes exposés sur la surface et encodé cette information dans des vecteurs. Nous avons utilisé l’analyse des composantes principales pour transformer les vecteurs en un ensemble de variables non corrélées, qu’on appelle les composantes principales. En utilisant la distance Euclidienne pondérée et l’algorithme du plus proche voisin, nous avons appliqué la technique du « Leave-One-Out Cross-Validation » pour choisir les meilleurs paramètres pour prédire l’activité d’une nouvelle séquence en la faisant correspondre à ces composantes principales. Finalement, nous avons confirmé le pouvoir prédictif du modèle à l’aide d’un nouvel ensemble de 8 variants dont la viabilité à été vérifiée expérimentalement dans notre laboratoire. En conclusion, les grammaires de graphes permettent de modéliser la relation entre la séquence et la structure d’un élément structural d’ARN, comme la boucle E contenant le motif Sarcin-Ricin du ribosome. Les applications vont de la correction à l’aide à l'alignement de séquences jusqu’au design de séquences ayant une structure prédéterminée. Nous avons également développé un modèle pour tenir compte des interactions spécifiques liées à une fonction biologique donnée, soit avec des facteurs environnants. Notre modèle est basé sur la conservation de l'exposition des groupes chimiques qui sont impliqués dans ces interactions. Ce modèle nous a permis de prédire l’activité biologique d’un ensemble de variants de la boucle E du ribosome qui se lie à des facteurs d'élongation.