Multivariate linear regression methods are widely used statistical tools in data analysis, and were developed when some response variables are studied simultaneously, in which our aim is to study the relationship between predictor variables and response variables through the regression coefficient matrix. The rapid improvements of information technology have brought us a large number of large-scale data, but also brought us great challenges in data processing. When dealing with high dimensional data, the classical least squares estimation is not applicable in multivariate linear regression analysis. In recent years, some approaches have been developed to deal with high-dimensional data problems, among which dimension reduction is one of the main approaches. In some literature, random projection methods were used to reduce dimension in large datasets. In Chapter 2, a new random projection method, with low-rank matrix approximation, is proposed to reduce the dimension of the parameter space in high-dimensional multivariate linear regression model. Some statistical properties of the proposed method are studied and explicit expressions are then derived for the accuracy loss of the method with Gaussian random projection and orthogonal random projection. These expressions are precise rather than being bounds up to constants. In multivariate regression analysis, reduced rank regression is also a dimension reduction method, which has become an important tool for achieving dimension reduction goals due to its simplicity, computational efficiency and good predictive performance. In practical situations, however, the performance of the reduced rank estimator is not satisfactory when the predictor variables are highly correlated or the ratio of signal to noise is small. To overcome this problem, in Chapter 3, we incorporate matrix projections into reduced rank regression method, and then develop reduced rank regression estimators based on random projection and orthogonal projection in high-dimensional multivariate linear regression models. We also propose a consistent estimator of the rank of the coefficient matrix and achieve prediction performance bounds for the proposed estimators based on mean squared errors. Envelope technology is also a popular method in recent years to reduce estimative and predictive variations in multivariate regression, including a class of methods to improve the efficiency without changing the traditional objectives. Variable selection is the process of selecting a subset of relevant features variables for use in model construction. The purpose of using this technology is to avoid the curse of dimensionality, simplify models to make them easier to interpret, shorten training time and reduce overfitting. In Chapter 4, we combine envelope models and a group variable selection method to propose an envelope-based sparse reduced rank regression estimator in high-dimensional multivariate linear regression models, and then establish its consistency, asymptotic normality and oracle property. Tensor data are in frequent use today in a variety of fields in science and engineering. Processing tensor data is a practical but challenging problem. Recently, the prevalence of tensor data has resulted in several envelope tensor versions. In Chapter 5, we incorporate envelope technique into tensor regression analysis and propose a partial tensor envelope model, which leads to a parsimonious version for tensor response regression when some predictors are of special interest, and then consistency and asymptotic normality of the coefficient estimators are proved. The proposed method achieves significant gains in efficiency compared to the standard tensor response regression model in terms of the estimation of the coefficients for the selected predictors. Finally, in Chapter 6, we summarize the work carried out in the thesis, and then suggest some problems of further research interest. / Dissertation / Doctor of Philosophy (PhD)

No description available.

<p dir="ltr">Since graphs are natural representations for encoding relational data, the problem of graph matching is an emerging task and has attracted increasing attention, which could potentially impact various domains such as social network de-anonymization and computer vision. Our main interest is designing polynomial-time algorithms for seeded graph matching problems where a subset of pre-matched vertex-pairs (seeds) is revealed. </p><p dir="ltr">However, the existing work does not fully investigate the pivotal role of seeds and falls short of making the most use of the seeds. Notably, the majority of existing hand-crafted algorithms only focus on using ``witnesses'' in the 1-hop neighborhood. Although some advanced algorithms are proposed to use multi-hop witnesses, their theoretical analysis applies only to \ER random graphs and requires seeds to be all correct, which often do not hold in real applications. Furthermore, a parallel line of research, Graph Neural Network (GNN) approaches, typically employs a semi-supervised approach, which requires a large number of seeds and lacks the capacity to distill knowledge transferable to unseen graphs.</p><p dir="ltr">In my dissertation, I have taken two approaches to address these limitations. In the first approach, we study to design hand-crafted algorithms that can properly use multi-hop witnesses to match graphs. We first study graph matching using multi-hop neighborhoods when partially-correct seeds are provided. Specifically, consider two correlated graphs whose edges are sampled independently from a parent \ER graph $\mathcal{G}(n,p)$. A mapping between the vertices of the two graphs is provided as seeds, of which an unknown fraction is correct. We first analyze a simple algorithm that matches vertices based on the number of common seeds in the $1$-hop neighborhoods, and then further propose a new algorithm that uses seeds in the $D$-hop neighborhoods. We establish non-asymptotic performance guarantees of perfect matching for both $1$-hop and $2$-hop algorithms, showing that our new $2$-hop algorithm requires substantially fewer correct seeds than the $1$-hop algorithm when graphs are sparse. Moreover, by combining our new performance guarantees for the $1$-hop and $2$-hop algorithms, we attain the best-known results (in terms of the required fraction of correct seeds) across the entire range of graph sparsity and significantly improve the previous results. We then study the role of multi-hop neighborhoods in matching power-law graphs. Assume that two edge-correlated graphs are independently edge-sampled from a common parent graph with a power-law degree distribution. A set of correctly matched vertex-pairs is chosen at random and revealed as initial seeds. Our goal is to use the seeds to recover the remaining latent vertex correspondence between the two graphs. Departing from the existing approaches that focus on the use of high-degree seeds in $1$-hop neighborhoods, we develop an efficient algorithm that exploits the low-degree seeds in suitably-defined $D$-hop neighborhoods. Our result achieves an exponential reduction in the seed size requirement compared to the best previously known results.</p><p dir="ltr">In the second approach, we study GNNs for seeded graph matching. We propose a new supervised approach that can learn from a training set how to match unseen graphs with only a few seeds. Our SeedGNN architecture incorporates several novel designs, inspired by our theoretical studies of seeded graph matching: 1) it can learn to compute and use witness-like information from different hops, in a way that can be generalized to graphs of different sizes; 2) it can use easily-matched node-pairs as new seeds to improve the matching in subsequent layers. We evaluate SeedGNN on synthetic and real-world graphs and demonstrate significant performance improvements over both non-learning and learning algorithms in the existing literature. Furthermore, our experiments confirm that the knowledge learned by SeedGNN from training graphs can be generalized to test graphs of different sizes and categories.</p>

In den vergangenen Jahren ist die Untersuchung des Risikomanagements vom Baselkomitee angeregt, um die Kredit- und Bankwesen regelmäßig zu aufsichten. Für viele multivariate Risikomanagementmethoden gibt es jedoch Beschränkungen von: 1) verlässt sich die Kovarianzschätzung auf eine zeitunabhängige Form, 2) die Modelle beruhen auf eine unrealistischen Verteilungsannahme und 3) numerische Problem, die bei hochdimensionalen Daten auftreten. Es ist das primäre Ziel dieser Doktorarbeit, präzise und schnelle Methoden vorzuschlagen, die diesen Beschränkungen überwinden. Die Grundidee besteht darin, zuerst aus einer hochdimensionalen Zeitreihe die stochastisch unabhängigen Komponenten (IC) zu extrahieren und dann die Verteilungsparameter der resultierenden IC beruhend auf eindimensionale Heavy-Tailed Verteilungsannahme zu identifizieren. Genauer gesagt werden zwei lokale parametrische Methoden verwendet, um den Varianzprozess jeder IC zu schätzen, das lokale Moving Window Average (MVA) Methode und das lokale Exponential Smoothing (ES) Methode. Diese Schätzungen beruhen auf der realistischen Annahme, dass die IC Generalized Hyperbolic (GH) verteilt sind. Die Berechnung ist schneller und erreicht eine höhere Genauigkeit als viele bekannte Risikomanagementmethoden. / Over recent years, study on risk management has been prompted by the Basel committee for the requirement of regular banking supervisory. There are however limitations of many risk management methods: 1) covariance estimation relies on a time-invariant form, 2) models are based on unrealistic distributional assumption and 3) numerical problems appear when applied to high-dimensional portfolios. The primary aim of this dissertation is to propose adaptive methods that overcome these limitations and can accurately and fast measure risk exposures of multivariate portfolios. The basic idea is to first retrieve out of high-dimensional time series stochastically independent components (ICs) and then identify the distributional behavior of every resulting IC in univariate space. To be more specific, two local parametric approaches, local moving window average (MWA) method and local exponential smoothing (ES) method, are used to estimate the volatility process of every IC under the heavy-tailed distributional assumption, namely ICs are generalized hyperbolic (GH) distributed. By doing so, it speeds up the computation of risk measures and achieves much better accuracy than many popular risk management methods.

Nous vivons à une époque marquée par une abondance de données cosmologiques de haute résolution. Cet afflux de données engendré par les missions d'observation de nouvelle génération au sol et dans l'espace porte le potentiel de remodeler fondamentalement notre compréhension de l'univers et de ses principes physiques sous-jacents. Cependant, la complexité grande des données observées pose des défis aux approches conventionnelles d'analyse de données, soit en raison de coûts de calcul irréalisables, soit en raison des hypothèses simplificatrices utilisées dans ces algorithmes qui deviennent inadéquates dans des contextes haute résolution à faible bruit, conduisant à des résultats sous-optimaux. En réponse, la communauté scientifique s'est tournée vers des méthodes innovantes d'analyse de données, notamment les techniques d'apprentissage automatique (ML). Les modèles de ML, lorsqu'ils sont bien entraînés, peuvent identifier de manière autonome des correlations significatives dans les données de manière plus efficace et sans hypothèses restrictives inutiles. Bien que les méthodes de ML aient montré des promesses en astrophysique, elles présentent également des problèmes tels que le manque d'interprétabilité, les biais cachés et les estimations d'incertitude non calibrées, ce qui, jusqu'a maintenant, a entrave leur application dans d'importantes découvertes scientifiques. Ce projet s'inscrit dans le cadre de la collaboration "Learning the Universe" (LtU), axée sur la reconstruction des conditions initiales de l'univers, en utilisant une approche de modélisation bayésienne et en exploitant la puissance du ML. L'objectif de ce projet est de développer un cadre pour mener une inférence bayésienne au niveau des pixels dans des problèmes multidimensionnels. Dans cette thèse, je présente le développement d'un cadre d'apprentissage profond pour un échantillonnage rapide des postérieurs en dimensions élevées. Ce cadre utilise l'architecture "Hierarchical Probabilistic U-Net", qui combine la puissance de l'architecture U-Net dans l'apprentissage de cartes multidimensionnelles avec le rigoureux cadre d'inférence des autoencodeurs variationnels conditionnels. Notre modèle peut quantifier les incertitudes dans ses données d'entraînement et générer des échantillons à partir de la distribution a posteriori des paramètres, pouvant être utilisés pour dériver des estimations d'incertitude pour les paramètres inférés. L'efficacité de notre cadre est démontrée en l'appliquant au problème de la reconstruction de cartes du fond diffus cosmologique (CMB) pour en retirer de l'effet de lentille gravitationnelle faible. Notre travail constitue un atout essentiel pour effectuer une inférence de vraisemblance implicite en dimensions élevées dans les domaines astrophysiques. Il permet d'exploiter pleinement le potentiel des missions d'observation de nouvelle génération pour améliorer notre compréhension de l'univers et de ses lois physiques fondamentales. / We live in an era marked by an abundance of high-resolution cosmological data. This influx of data brought about by next-generation observational missions on the ground and in space, bears the potential of fundamentally reshaping our understanding of the universe and its underlying physical principles. However, the elevated complexity of the observed data poses challenges to conventional data analysis approaches, either due to infeasible computational costs or the simplifying assumptions used in these algorithms that become inadequate in high-resolution, low-noise contexts, leading to suboptimal results. In response, the scientific community has turned to innovative data analysis methods, including machine learning (ML) techniques. ML models, when well-trained, can autonomously identify meaningful patterns in data more efficiently and without unnecessary restrictive assumptions. Although ML methods have shown promise in astrophysics, they also exhibit issues like lack of interpretability, hidden biases, and uncalibrated uncertainty estimates, which have hindered their application in significant scientific discoveries. This project is defined within the context of the Learning the Universe (LtU) collaboration, focused on reconstructing the initial conditions of the universe, utilizing a Bayesian forward modeling approach and harnessing the power of ML. The goal of this project is to develop a framework for conducting Bayesian inference at the pixel level in high-dimensional problems. In this thesis, I present the development of a deep learning framework for fast high-dimensional posterior sampling. This framework utilizes the Hierarchical Probabilistic U-Net architecture, which combines the power of the U-Net architecture in learning high-dimensional mappings with the rigorous inference framework of Conditional Variational Autoencoders. Our model can quantify uncertainties in its training data and generate samples from the posterior distribution of parameters, which can be used to derive uncertainty estimates for the inferred parameters. The effectiveness of our framework is demonstrated by applying it to the problem of removing the weak gravitational lensing effect from the CMB. Our work stands as an essential asset to performing high-dimensional implicit likelihood inference in astrophysical domains. It enables utilizing the full potential of next-generation observational missions to improve our understanding of the universe and its fundamental physical laws.

No description available.

A Growth Curve Model (GCM) is a multivariate linear model used for analyzing longitudinal data with short to moderate time series. It is a special case of Generalized Multivariate Analysis of Variance (GMANOVA) models. Analysis using the GCM involves comparison of mean growths among different groups. The classical GCM, however, possesses some limitations including distributional assumptions, assumption of identical degree of polynomials for all groups and it requires larger sample size than the number of time points. In this thesis, we relax some of the assumptions of the traditional GCM and develop appropriate inferential tools for its analysis, with the aim of reducing bias, improving precision and to gain increased power as well as overcome limitations of high-dimensionality. Existing methods for estimating the parameters of the GCM assume that the underlying distribution for the error terms is multivariate normal. In practical problems, however, we often come across skewed data and hence estimation techniques developed under the normality assumption may not be optimal. Simulation studies conducted in this thesis, in fact, show that existing methods are sensitive to the presence of skewness in the data, where estimators are associated with increased bias and mean square error (MSE), when the normality assumption is violated. Methods appropriate for skewed distributions are, therefore, required. In this thesis, we relax the distributional assumption of the GCM and provide estimators for the mean and covariance matrices of the GCM under multivariate skew normal (MSN) distribution. An estimator for the additional skewness parameter of the MSN distribution is also provided. The estimators are derived using the expectation maximization (EM) algorithm and extensive simulations are performed to examine the performance of the estimators. Comparisons with existing estimators show that our estimators perform better than existing estimators, when the underlying distribution is multivariate skew normal. Illustration using real data set is also provided, wherein Triglyceride levels from the Framingham Heart Study is modelled over time. The GCM assumes equal degree of polynomial for each group. Therefore, when groups means follow different shapes of polynomials, the GCM fails to accommodate this difference in one model. We consider an extension of the GCM, wherein mean responses from different groups can have different shapes, represented by polynomials of different degree. Such a model is referred to as Extended Growth Curve Model (EGCM). We extend our work on GCM to EGCM, and develop estimators for the mean and covariance matrices under MSN errors. We adopted the Restricted Expectation Maximization (REM) algorithm, which is based on the multivariate Newton-Raphson (NR) method and Lagrangian optimization. However, the multivariate NR method and hence, the existing REM algorithm are applicable to vector parameters and the parameters of interest in this study are matrices. We, therefore, extended the NR approach to matrix parameters, which consequently allowed us to extend the REM algorithm to matrix parameters. The performance of the proposed estimators were examined using extensive simulations and a motivating real data example was provided to illustrate the application of the proposed estimators. Finally, this thesis deals with high-dimensional application of GCM. Existing methods for a GCM are developed under the assumption of ‘small p large n’ (n >> p) and are not appropriate for analyzing high-dimensional longitudinal data, due to singularity of the sample covariance matrix. In a previous work, we used Moore-Penrose generalized inverse to overcome this challenge. However, the method has some limitations around near singularity, when p~n. In this thesis, a Bayesian framework was used to derive a test for testing the linear hypothesis on the mean parameter of the GCM, which is applicable in high-dimensional situations. Extensive simulations are performed to investigate the performance of the test statistic and establish optimality characteristics. Results show that this test performs well, under different conditions, including the near singularity zone. Sensitivity of the test to mis-specification of the parameters of the prior distribution are also examined empirically. A numerical example is provided to illustrate the usefulness of the proposed method in practical situations. / Thesis / Doctor of Philosophy (PhD)

<p>The need to carry out Uncertainty quantification (UQ) is ubiquitous in science and engineering. However, carrying out UQ for real-world problems is not straightforward and they require a lot of computational budget and resources. The objective of this thesis is to develop computationally efficient approaches based on machine learning to carry out UQ. Specifically, we addressed two problems.</p> <p><br></p> <p>The first problem is, it is difficult to carry out Uncertainty propagation (UP) in systems governed by elliptic PDEs with spatially varying uncertain fields in coefficients and boundary conditions. Here as we have functional uncertainties, the number of uncertain parameters is large. Unfortunately, in these situations to carry out UP we need to solve the PDE a large number of times to obtain convergent statistics of the quantity governed by the PDE. However, solving the PDE by a numerical solver repeatedly leads to a computational burden. To address this we proposed to learn the surrogate of the solution of the PDE in a data-free manner by utilizing the physics available in the form of the PDE. We represented the solution of the PDE as a deep neural network parameterized function in space and uncertain parameters. We introduced a physics-informed loss function derived from variational principles to learn the parameters of the network. The accuracy of the learned surrogate is validated against the corresponding ground truth estimate from the numerical solver. We demonstrated the merit of using our approach by solving UP problems and inverse problems faster than by using a standard numerical solver.</p> <p><br></p> <p>The second problem we focused on in this thesis is related to inverse problems. State of the art approach to solving inverse problems involves posing the inverse problem as a Bayesian inference task and estimating the distribution of input parameters conditioned on the observed data (posterior). Markov Chain Monte Carlo (MCMC) methods and variational inference methods provides us ways to estimate the posterior. However, these inference techniques need to be re-run whenever a new set of observed data is given leading to a computational burden. To address this, we proposed to learn a Bayesian inverse map i.e., the map from the observed data to the posterior. This map enables us to do on-the-fly inference. We demonstrated our approach by solving various examples and we validated the posteriors learned from our approach against corresponding ground truth posteriors from the MCMC method.</p>

En finance, les modèles d’évaluation des actifs tentent de comprendre les différences de rendements observées entre divers actifs. Hansen and Richard (1987) ont montré que ces modèles sont des représentations fonctionnelles du facteur d’actualisation stochastique que les investisseurs utilisent pour déterminer le prix des actifs sur le marché financier. La littérature compte de nombreuses études économétriques qui s’intéressent à leurs estimations et à la comparaison de leurs performances, c’est-à-dire de leur capa- cité à expliquer les différences de rendement observées. Cette thèse, composée de trois articles, contribue à cette littérature. Le premier article examine l’estimation et la comparaison des modèles d’évaluation des actifs dans un environnement riche en données. Nous mettons en œuvre deux méthodes de régularisation interprétables de la distance de Hansen and Jagannathan (1997, HJ ci-après) dans un contexte où les actifs sont nombreux. Plus précisément, nous introduisons la régularisation de Tikhonov et de Ridge pour stabiliser l’inverse de la matrice de covariance de la distance de HJ. La nouvelle mesure, qui en résulte, peut être interprétée comme la distance entre le facteur d’actualisation d’un modèle et le facteur d’actualisation stochastique valide le plus proche qui évalue les actifs avec des erreurs contrôlées. Ainsi, ces méthodes de régularisation relâchent l’équation fondamentale de l’évaluation des actifs financiers. Aussi, elles incorporent un paramètre de régularisation régissant l’ampleur des erreurs d’évaluation. Par la suite, nous présentons une procédure pour estimer et faire des tests sur les paramètres d’un modèle d’évaluation des actifs financiers avec un facteur d’actualisation linéaire en minimisant la distance de HJ régularisée. De plus, nous obtenons la distribution asymptotique des estimateurs lorsque le nombre d’actifs devient grand. Enfin, nous déterminons la distribution de la distance régularisée pour comparer différents modèles d’évaluation des actifs. Empiriquement, nous estimons et comparons quatre modèles à l’aide d’un ensemble de données comportant 252 portefeuilles. Le deuxième article estime et compare dix modèles d’évaluation des actifs, à la fois inconditionnels et conditionnels, en utilisant la distance de HJ régularisée et 3 198 portefeuilles s’étendant de juillet 1973 à juin 2018. Ces portefeuilles combinent les portefeuilles bien connus triés par caractéristiques avec des micro-portefeuilles. Les micro-portefeuilles sont formés à l’aide de variables financières mais contiennent peu d’actions (5 à 10), comme indiqué dans Barras (2019). Par conséquent, ils sont analogues aux actions individuelles, offrent une grande variabilité de rendements et améliorent le pouvoir discriminant des portefeuilles classiques triés par caractéristiques. Parmi les modèles considérés, quatre sont des modèles macroéconomiques ou théoriques, dont le modèle de CAPM avec consommation (CCAPM), le modèle de CAPM avec consommation durable (DCAPM) de Yogo (2006), le modèle de CAPM avec capital humain (HCAPM) de Jagannathan and Wang (1996), et le modèle d’évaluation des actifs avec intermédiaires financiers (IAPM) de He, Kelly, and Manela (2017). Cinq modèles basés sur les anomalies sont considérés, tels que les modèles à trois (FF3) et à cinq facteurs (FF5) proposés par Fama and French, 1993 et 2015, le modèle de Carhart (1997) intégrant le facteur Momentum dans FF3, le modèle de liquidité de Pástor and Stambaugh (2003) et le modèle q5 de Hou et al. (2021). Le modèle de consommation de Lettau and Ludvigson (2001) utilisant des données trimestrielles est également estimé. Cependant, il n’est pas inclus dans les comparaisons en raison de la puissance de test réduite. Par rapport aux modèles inconditionnels, les modèles conditionnels tiennent compte des cycles économiques et des fluctuations des marchés financiers en utilisant les indices d’incertitude macroéconomique et financière de Ludvigson, Ma, and Ng (2021). Ces modèles conditionnels ont des erreurs de spécification considérablement réduites. Les analyses comparatives des modèles inconditionnels indiquent que les modèles macroéconomiques présentent globalement les mêmes pouvoirs explicatifs. De plus, ils ont un pouvoir explicatif global inférieur à celui des modèles basés sur les anomalies, à l’exception de FF3. L’augmentation de FF3 avec le facteur Momentum et de liquidité améliore sa capacité explicative. Cependant ce nouveau modèle est inférieur à FF5 et q5. Pour les modèles conditionnels, les modèles macroéconomiques DCAPM et HCAPM surpassent CCAPM et IAPM. En outre, ils ont des erreurs de spécification similaires à celles des modèles conditionnels de Carhart et de liquidité, mais restent en deçà des modèles FF5 et q5. Ce dernier domine tous les autres modèles. Le troisième article présente une nouvelle approche pour estimer les paramètres du facteur d’actualisation linéaire des modèles d’évaluation d’actifs linéaires mal spécifiés avec de nombreux actifs. Contrairement au premier article de Carrasco and Nokho (2022), cette approche s’applique à la fois aux rendements bruts et excédentaires. La méthode proposée régularise toujours la distance HJ : l’inverse de la matrice de second moment est la matrice de pondération pour les rendements bruts, tandis que pour les rendements excédentaires, c’est l’inverse de la matrice de covariance. Plus précisément, nous dérivons la distribution asymptotique des estimateurs des paramètres du facteur d’actualisation stochastique lorsque le nombre d’actifs augmente. Nous discutons également des considérations pertinentes pour chaque type de rendements et documentons les propriétés d’échantillon fini des estimateurs. Nous constatons qu’à mesure que le nombre d’actifs augmente, l’estimation des paramètres par la régularisation de l’inverse de la matrice de covariance des rendements excédentaires présente un contrôle de taille supérieur par rapport à la régularisation de l’inverse de la matrice de second moment des rendements bruts. Cette supériorité découle de l’instabilité inhérente à la matrice de second moment des rendements bruts. De plus, le rendement brut de l’actif sans risque présente une variabilité minime, ce qui entraîne une colinéarité significative avec d’autres actifs que la régularisation ne parvient pas à atténuer. / In finance, asset pricing models try to understand the differences in expected returns observed among various assets. Hansen and Richard (1987) showed that these models are functional representations of the discount factor investors use to price assets in the financial market. The literature counts many econometric studies that deal with their estimation and the comparison of their performance, i.e., how well they explain the differences in expected returns. This thesis, divided into three chapters, contributes to this literature. The first paper examines the estimation and comparison of asset pricing models in a data-rich environment. We implement two interpretable regularization schemes to extend the renowned Hansen and Jagannathan (1997, HJ hereafter) distance to a setting with many test assets. Specifically, we introduce Tikhonov and Ridge regularizations to stabilize the inverse of the covariance matrix in the HJ distance. The resulting misspecification measure can be interpreted as the distance between a proposed pricing kernel and the nearest valid stochastic discount factor (SDF) pricing the test assets with controlled errors, relaxing the Fundamental Equation of Asset Pricing. So, these methods incorporate a regularization parameter governing the extent of the pricing errors. Subsequently, we present a procedure to estimate the SDF parameters of a linear asset pricing model by minimizing the regularized distance. The SDF parameters completely define the asset pricing model and determine if a particular observed factor is a priced source of risk in the test assets. In addition, we derive the asymptotic distribution of the estimators when the number of assets and time periods increases. Finally, we derive the distribution of the regularized distance to compare comprehensively different asset pricing models. Empirically, we estimate and compare four empirical asset pricing models using a dataset of 252 portfolios. The second paper estimates and compares ten asset pricing models, both unconditional and conditional, utilizing the regularized HJ distance and 3198 portfolios spanning July 1973 to June 2018. These portfolios combine the well-known characteristic-sorted portfolios with micro portfolios. The micro portfolios are formed using firms' observed financial characteristics (e.g. size and book-to-market) but contain few stocks (5 to 10), as discussed in Barras (2019). Consequently, they are analogous to individual stocks, offer significant return spread, and improve the discriminatory power of the characteristics-sorted portfolios. Among the models, four are macroeconomic or theoretical models, including the Consumption Capital Asset Pricing Model (CCAPM), Durable Consumption Capital Asset Pricing Model (DCAPM) by Yogo (2006), Human Capital Capital Asset Pricing Model (HCAPM) by Jagannathan and Wang (1996), and Intermediary Asset pricing model (IAPM) by He, Kelly, and Manela (2017). Five anomaly-driven models are considered, such as the three (FF3) and Five-factor (FF5) Models proposed by Fama and French, 1993 and 2015, the Carhart (1997) model incorporating momentum into FF3, the Liquidity Model by Pástor and Stambaugh (2003), and the Augmented q-Factor Model (q5) by Hou et al. (2021). The Consumption model of Lettau and Ludvigson (2001) using quarterly data is also estimated but not included in the comparisons due to the reduced power of the tests. Compared to the unconditional models, the conditional ones account for the economic business cycles and financial market fluctuations by utilizing the macroeconomic and financial uncertainty indices of Ludvigson, Ma, and Ng (2021). These conditional models show significantly reduced pricing errors. Comparative analyses of the unconditional models indicate that the macroeconomic models exhibit similar pricing performances of the returns. In addition, they display lower overall explanatory power than anomaly-driven models, except for FF3. Augmenting FF3 with momentum and liquidity factors enhances its explanatory capability. However, the new model is inferior to FF5 and q5. For the conditional models, the macroeconomic models DCAPM and HCAPM outperform CCAPM and IAPM. Furthermore, they have similar pricing errors as the conditional Carhart and liquidity models but still fall short of the FF5 and q5. The latter dominates all the other models. This third paper introduces a novel approach for estimating the SDF parameters in misspecified linear asset pricing models with many assets. Unlike the first paper, Carrasco and Nokho (2022), this approach is applicable to both gross and excess returns as test assets. The proposed method still regularizes the HJ distance: the inverse of the second-moment matrix is the weighting matrix for the gross returns, while for excess returns, it is the inverse of the covariance matrix. Specifically, we derive the asymptotic distribution of the SDF estimators under a double asymptotic condition where the number of test assets and time periods go to infinity. We also discuss relevant considerations for each type of return and document the finite sample properties of the SDF estimators with gross and excess returns. We find that as the number of test assets increases, the estimation of the SDF parameters through the regularization of the inverse of the excess returns covariance matrix exhibits superior size control compared to the regularization of the inverse of the gross returns second-moment matrix. This superiority arises from the inherent instability of the second-moment matrix of gross returns. Additionally, the gross return of the risk-free asset shows minimal variability, resulting in significant collinearity with other test assets that the regularization fails to mitigate.

We consider multivariate trigonometric polynomials with frequencies supported on a fixed but arbitrary frequency index set I, which is a finite set of integer vectors of length d. Naturally, one is interested in spatial discretizations in the d-dimensional torus such that - the sampling values of the trigonometric polynomial at the nodes of this spatial discretization uniquely determines the trigonometric polynomial, - the corresponding discrete Fourier transform is fast realizable, and - the corresponding fast Fourier transform is stable. An algorithm that computes the discrete Fourier transform and that needs a computational complexity that is bounded from above by terms that are linear in the maximum of the number of input and output data up to some logarithmic factors is called fast Fourier transform. We call the fast Fourier transform stable if the Fourier matrix of the discrete Fourier transform has a condition number near one and the fast algorithm does not corrupt this theoretical stability. We suggest to use rank-1 lattices and a generalization as spatial discretizations in order to sample multivariate trigonometric polynomials and we develop construction methods in order to determine reconstructing sampling sets, i.e., sets of sampling nodes that allow for the unique, fast, and stable reconstruction of trigonometric polynomials. The methods for determining reconstructing rank-1 lattices are component{by{component constructions, similar to the seminal methods that are developed in the field of numerical integration. During this thesis we identify a component{by{component construction of reconstructing rank-1 lattices that allows for an estimate of the number of sampling nodes M |I|\le M\le \max\left(\frac{2}{3}|I|^2,\max\{3\|\mathbf{k}\|_\infty\colon\mathbf{k}\in I\}\right) that is sufficient in order to uniquely reconstruct each multivariate trigonometric polynomial with frequencies supported on the frequency index set I. We observe that the bounds on the number M only depends on the number of frequency indices contained in I and the expansion of I, but not on the spatial dimension d. Hence, rank-1 lattices are suitable spatial discretizations in arbitrarily high dimensional problems. Furthermore, we consider a generalization of the concept of rank-1 lattices, which we call generated sets. We use a quite different approach in order to determine suitable reconstructing generated sets. The corresponding construction method is based on a continuous optimization method. Besides the theoretical considerations, we focus on the practicability of the presented algorithms and illustrate the theoretical findings by means of several examples. In addition, we investigate the approximation properties of the considered sampling schemes. We apply the results to the most important structures of frequency indices in higher dimensions, so-called hyperbolic crosses and demonstrate the approximation properties by the means of several examples that include the solution of Poisson's equation as one representative of partial differential equations.