arXiv subjects

Marc G. Genton

Publications and source records attributed to Marc G. Genton.

At least 73 records · Page 4Linked to original sources

arXiv2020

Trajectory Functional Boxplots

With the development of data-monitoring techniques in various fields of science, multivariate functional data are often observed. Consequently, an increasing number of methods have appeared to extend the general summary statistics of multivariate functional data. However, trajectory functional data, as an important sub-type, have not been studied very well. This article proposes two informative exploratory tools, the trajectory functional boxplot, and the modified simplicial band depth (MSBD) versus Wiggliness of Directional Outlyingness (WO) plot, to visualize the centrality of trajectory functional data. The newly defined WO index effectively measures the shape variation of curves and hence serves as a detector for shape outliers; additionally, MSBD provides a center-outward ranking result and works as a detector for magnitude outliers. Using the two measures, the functional boxplot of the trajectory reveals center-outward patterns and potential outliers using the raw curves, whereas the MSBD-WO plot illustrates such patterns and outliers in a space spanned by MSBD and WO. The proposed methods are validated on hurricane path data and migration trace data recorded from two types of birds.

stat.ME↗

arXiv2020

A Generalized Heckman Model With Varying Sample Selection Bias and Dispersion Parameters

Many proposals have emerged as alternatives to the Heckman selection model, mainly to address the non-robustness of its normal assumption. The 2001 Medical Expenditure Panel Survey data is often used to illustrate this non-robustness of the Heckman model. In this paper, we propose a generalization of the Heckman sample selection model by allowing the sample selection bias and dispersion parameters to depend on covariates. We show that the non-robustness of the Heckman model may be due to the assumption of the constant sample selection bias parameter rather than the normality assumption. Our proposed methodology allows us to understand which covariates are important to explain the sample selection bias phenomenon rather than to only form conclusions about its presence. We explore the inferential aspects of the maximum likelihood estimators (MLEs) for our proposed generalized Heckman model. More specifically, we show that this model satisfies some regularity conditions such that it ensures consistency and asymptotic normality of the MLEs. Proper score residuals for sample selection models are provided, and model adequacy is addressed. Simulated results are presented to check the finite-sample behavior of the estimators and to verify the consequences of not considering varying sample selection bias and dispersion parameters. We show that the normal assumption for analyzing medical expenditure data is suitable and that the conclusions drawn using our approach are coherent with findings from prior literature. Moreover, we identify which covariates are relevant to explain the presence of sample selection bias in this important dataset.

stat.ME↗

arXiv2020

Exploiting Low Rank Covariance Structures for Computing High-Dimensional Normal and Student-$t$ Probabilities

We present a preconditioned Monte Carlo method for computing high-dimensional multivariate normal and Student-$t$ probabilities arising in spatial statistics. The approach combines a tile-low-rank representation of covariance matrices with a block-reordering scheme for efficient Quasi-Monte Carlo simulation. The tile-low-rank representation decomposes the high-dimensional problem into many diagonal-block-size problems and low-rank connections. The block-reordering scheme reorders between and within the diagonal blocks to reduce the impact of integration variables from right to left, thus improving the Monte Carlo convergence rate. Simulations up to dimension $65{,}536$ suggest that the new method can improve the run time by an order of magnitude compared with the non-reordered tile-low-rank Quasi-Monte Carlo method and two orders of magnitude compared with the dense Quasi-Monte Carlo method. Our method also forms a strong substitute for the approximate conditioning methods as a more robust estimation with error guarantees. An application study to wind stochastic generators is provided to illustrate that the new computational method makes the maximum likelihood estimation feasible for high-dimensional skew-normal random fields.

stat.CO↗

arXiv2020

Nonparametric Trend Estimation in Functional Time Series with Application to Annual Mortality Rates

Here, we address the problem of trend estimation for functional time series. Existing contributions either deal with detecting a functional trend or assuming a simple model. They consider neither the estimation of a general functional trend nor the analysis of functional time series with a functional trend component. Similarly to univariate time series, we propose an alternative methodology to analyze functional time series, taking into account a functional trend component. We propose to estimate the functional trend by using a tensor product surface that is easy to implement, to interpret, and allows to control the smoothness properties of the estimator. Through a Monte Carlo study, we simulate different scenarios of functional processes to show that our estimator accurately identifies the functional trend component. We also show that the dependency structure of the estimated stationary time series component is not significantly affected by the error approximation of the functional trend component. We apply our methodology to annual mortality rates in France.

stat.ME↗

arXiv2020

Improving Bayesian Local Spatial Models in Large Data Sets

Environmental processes resolved at a sufficiently small scale in space and time will inevitably display non-stationary behavior. Such processes are both challenging to model and computationally expensive when the data size is large. Instead of modeling the global non-stationarity explicitly, local models can be applied to disjoint regions of the domain. The choice of the size of these regions is dictated by a bias-variance trade-off; large regions will have smaller variance and larger bias, whereas small regions will have higher variance and smaller bias. From both the modeling and computational point of view, small regions are preferable to better accommodate the non-stationarity. However, in practice, large regions are necessary to control the variance. We propose a novel Bayesian three-step approach that allows for smaller regions without compromising the increase of the variance that would follow. We are able to propagate the uncertainty from one step to the next without issues caused by reusing the data. The improvement in inference also results in improved prediction, as our simulated example shows. We illustrate this new approach on a data set of simulated high-resolution wind speed data over Saudi Arabia.

stat.AP↗

arXiv2020

Recent Developments in Complex and Spatially Correlated Functional Data

As high-dimensional and high-frequency data are being collected on a large scale, the development of new statistical models is being pushed forward. Functional data analysis provides the required statistical methods to deal with large-scale and complex data by assuming that data are continuous functions, e.g., a realization of a continuous process (curves) or continuous random fields (surfaces), and that each curve or surface is considered as a single observation. Here, we provide an overview of functional data analysis when data are complex and spatially correlated. We provide definitions and estimators of the first and second moments of the corresponding functional random variable. We present two main approaches: The first assumes that data are realizations of a functional random field, i.e., each observation is a curve with a spatial component. We call them 'spatial functional data'. The second approach assumes that data are continuous deterministic fields observed over time. In this case, one observation is a surface or manifold, and we call them 'surface time series'. For the two approaches, we describe software available for the statistical analysis. We also present a data illustration, using a high-resolution wind speed simulated dataset, as an example of the two approaches. The functional data approach offers a new paradigm of data analysis, where the continuous processes or random fields are considered as a single entity. We consider this approach to be very valuable in the context of big data.

math.ST↗

arXiv2020

Functional Outlier Detection and Taxonomy by Sequential Transformations

Functional data analysis can be seriously impaired by abnormal observations, which can be classified as either magnitude or shape outliers based on their way of deviating from the bulk of data. Identifying magnitude outliers is relatively easy, while detecting shape outliers is much more challenging. We propose turning the shape outliers into magnitude outliers through data transformation and detecting them using the functional boxplot. Besides easing the detection procedure, applying several transformations sequentially provides a reasonable taxonomy for the flagged outliers. A joint functional ranking, which consists of several transformations, is also defined here. Simulation studies are carried out to evaluate the performance of the proposed method using different functional depth notions. Interesting results are obtained in several practical applications.

stat.ME↗

arXiv2020

A Pairwise Hotelling Method for Testing High-Dimensional Mean Vectors

For high-dimensional small sample size data, Hotelling's T2 test is not applicable for testing mean vectors due to the singularity problem in the sample covariance matrix. To overcome the problem, there are three main approaches in the literature. Note, however, that each of the existing approaches may have serious limitations and only works well in certain situations. Inspired by this, we propose a pairwise Hotelling method for testing high-dimensional mean vectors, which, in essence, provides a good balance between the existing approaches. To effectively utilize the correlation information, we construct the new test statistics as the summation of Hotelling's test statistics for the covariate pairs with strong correlations and the squared $t$ statistics for the individual covariates that have little correlation with others. We further derive the asymptotic null distributions and power functions for the proposed Hotelling tests under some regularity conditions. Numerical results show that our new tests are able to control the type I error rates, and can achieve a higher statistical power compared to existing methods, especially when the covariates are highly correlated. Two real data examples are also analyzed and they both demonstrate the efficacy of our pairwise Hotelling tests.

stat.ME↗

arXiv2020

Geostatistical Modeling and Prediction Using Mixed-Precision Tile Cholesky Factorization

Geostatistics represents one of the most challenging classes of scientific applications due to the desire to incorporate an ever increasing number of geospatial locations to accurately model and predict environmental phenomena. For example, the evaluation of the Gaussian log-likelihood function, which constitutes the main computational phase, involves solving systems of linear equations with a large dense symmetric and positive definite covariance matrix. Cholesky, the standard algorithm, requires O(n^3) floating point operators and has an O(n^2) memory footprint, where n is the number of geographical locations. Here, we present a mixed-precision tile algorithm to accelerate the Cholesky factorization during the log-likelihood function evaluation. Under an appropriate ordering, it operates with double-precision arithmetic on tiles around the diagonal, while reducing to single-precision arithmetic for tiles sufficiently far off. This translates into an improvement of the performance without any deterioration of the numerical accuracy of the application. We rely on the StarPU dynamic runtime system to schedule the tasks and to overlap them with data movement. To assess the performance and the accuracy of the proposed mixed-precision algorithm, we use synthetic and real datasets on various shared and distributed-memory systems possibly equipped with hardware accelerators. We compare our mixed-precision Cholesky factorization against the double-precision reference implementation as well as an independent block approximation method. We obtain an average of 1.6X performance speedup on massively parallel architectures while maintaining the accuracy necessary for modeling and prediction.

cs.DC↗

arXiv2019

Spatial Blind Source Separation

Recently a blind source separation model was suggested for spatial data together with an estimator based on the simultaneous diagonalisation of two scatter matrices. The asymptotic properties of this estimator are derived here and a new estimator, based on the joint diagonalisation of more than two scatter matrices, is proposed. The asymptotic properties and merits of the novel estimator are verified in simulation studies. A real data example illustrates the method.

math.ST↗

arXiv2019

A Non-Gaussian Spatio-Temporal Model for Daily Wind Speeds Based on a Multivariate Skew-t Distribution

Facing increasing domestic energy consumption from population growth and industrialization, Saudi Arabia is aiming to reduce its reliance on fossil fuels and to broaden its energy mix by expanding investment in renewable energy sources, including wind energy. A preliminary task in the development of wind energy infrastructure is the assessment of wind energy potential, a key aspect of which is the characterization of its spatio-temporal behavior. In this study we examine the impact of internal climate variability on seasonal wind power density fluctuations over Saudi Arabia using 30 simulations from the Large Ensemble Project (LENS) developed at the National Center for Atmospheric Research. Furthermore, a spatio-temporal model for daily wind speed is proposed with neighbor-based cross-temporal dependence, and a multivariate skew-t distribution to capture the spatial patterns of higher order moments. The model can be used to generate synthetic time series over the entire spatial domain that adequately reproduce the internal variability of the LENS dataset.

stat.AP↗

arXiv2018

A Copula Model for Non-Gaussian Multivariate Spatial Data

We propose a new copula model for replicated multivariate spatial data. Unlike classical models that assume multivariate normality of the data, the proposed copula is based on the assumption that some factors exist that affect the joint spatial dependence of all measurements of each variable as well as the joint dependence among these variables. The model is parameterized in terms of a cross-covariance function that may be chosen from the many models proposed in the literature. In addition, there are additive factors in the model that allow tail dependence and reflection asymmetry of each variable measured at different locations and of different variables to be modeled. The proposed approach can therefore be seen as an extension of the linear model of coregionalization widely used for modeling multivariate spatial data. The likelihood of the model can be obtained in a simple form and therefore the likelihood estimation is quite fast. The model is not restricted to the set of data locations, and using the estimated copula, spatial data can be interpolated at locations where values of variables are unknown. We apply the proposed model to temperature and pressure data and compare its performance with the performance of a popular model from multivariate geostatistics.

stat.AP↗

arXiv2018

Diagonal Likelihood Ratio Test for Equality of Mean Vectors in High-Dimensional Data

We propose a likelihood ratio test framework for testing normal mean vectors in high-dimensional data under two common scenarios: the one-sample test and the two-sample test with equal covariance matrices. We derive the test statistics under the assumption that the covariance matrices follow a diagonal matrix structure. In comparison with the diagonal Hotelling's tests, our proposed test statistics display some interesting characteristics. In particular, they are a summation of the log-transformed squared $t$-statistics rather than a direct summation of those components. More importantly, to derive the asymptotic normality of our test statistics under the null and local alternative hypotheses, we do not need the requirement that the covariance matrices follow a diagonal matrix structure. As a consequence, our proposed test methods are very flexible and readily applicable in practice. Simulation studies and a real data analysis are also carried out to demonstrate the advantages of our likelihood ratio test methods.

stat.ME↗

arXiv2018

O(N) Hierarchical algorithm for computing the expectations of truncated multi-variate normal distributions in N dimensions

In this paper, we study the $N$-dimensional integral $ϕ(a,b; A) = \int_{a}^{b} H(x) f(x | A) \text{d} x$ representing the expectation of a function $H(X)$ where $f(x | A)$ is the truncated multi-variate normal (TMVN) distribution with zero mean, $x$ is the vector of integration variables for the $N$-dimensional random vector $X$, $A$ is the inverse of the covariance matrix $Σ$, and $a$ and $b$ are constant vectors. We present a new hierarchical algorithm which can evaluate $ϕ(a,b; A)$ using asymptotically optimal $O(N)$ operations when $A$ has "low-rank" blocks with "low-dimensional" features and $H(x)$ is "low-rank". We demonstrate the divide-and-conquer idea when $A$ is a symmetric positive definite tridiagonal matrix, and present the necessary building blocks and rigorous potential theory based algorithm analysis when $A$ is given by the exponential covariance model. Numerical results are presented to demonstrate the algorithm accuracy and efficiency for these two cases. We also briefly discuss how the algorithm can be generalized to a wider class of covariance models and its limitations.

math.NA↗

arXiv2018

Likelihood Approximation With Hierarchical Matrices For Large Spatial Datasets

We use available measurements to estimate the unknown parameters (variance, smoothness parameter, and covariance length) of a covariance function by maximizing the joint Gaussian log-likelihood function. To overcome cubic complexity in the linear algebra, we approximate the discretized covariance function in the hierarchical (H-) matrix format. The H-matrix format has a log-linear computational cost and storage $O(kn \log n)$, where the rank $k$ is a small integer and $n$ is the number of locations. The H-matrix technique allows us to work with general covariance matrices in an efficient way, since H-matrices can approximate inhomogeneous covariance functions, with a fairly general mesh that is not necessarily axes-parallel, and neither the covariance matrix itself nor its inverse have to be sparse. We demonstrate our method with Monte Carlo simulations and an application to soil moisture data. The C, C++ codes and data are freely available.

stat.CO↗

arXiv2018

Bayesian model averaging over tree-based dependence structures for multivariate extremes

Describing the complex dependence structure of extreme phenomena is particularly challenging. To tackle this issue we develop a novel statistical algorithm that describes extremal dependence taking advantage of the inherent hierarchical dependence structure of the max-stable nested logistic distribution and that identifies possible clusters of extreme variables using reversible jump Markov chain Monte Carlo techniques. Parsimonious representations are achieved when clusters of extreme variables are found to be completely independent. Moreover, we significantly decrease the computational complexity of full likelihood inference by deriving a recursive formula for the nested logistic model likelihood. The algorithm performance is verified through extensive simulation experiments which also compare different likelihood procedures. The new methodology is used to investigate the dependence relationships between extreme concentration of multiple pollutants in California and how these pollutants are related to extreme weather conditions. Overall, we show that our approach allows for the representation of complex extremal dependence structures and has valid applications in multivariate data analysis, such as air pollution monitoring, where it can guide policymaking.

stat.ME↗

arXiv2018

Full likelihood inference for max-stable data

We show how to perform full likelihood inference for max-stable multivariate distributions or processes based on a stochastic Expectation-Maximisation algorithm, which combines statistical and computational efficiency in high-dimensions. The good performance of this methodology is demonstrated by simulation based on the popular logistic and Brown--Resnick models, and it is shown to provide dramatic computational time improvements with respect to a direct computation of the likelihood. Strategies to further reduce the computational burden are also discussed.

stat.ME↗

arXiv2018

ExaGeoStat: A High Performance Unified Software for Geostatistics on Manycore Systems

We present ExaGeoStat, a high performance framework for geospatial statistics in climate and environment modeling. In contrast to simulation based on partial differential equations derived from first-principles modeling, ExaGeoStat employs a statistical model based on the evaluation of the Gaussian log-likelihood function, which operates on a large dense covariance matrix. Generated by the parametrizable Matern covariance function, the resulting matrix is symmetric and positive definite. The computational tasks involved during the evaluation of the Gaussian log-likelihood function become daunting as the number n of geographical locations grows, as O(n2) storage and O(n3) operations are required. While many approximation methods have been devised from the side of statistical modeling to ameliorate these polynomial complexities, we are interested here in the complementary approach of evaluating the exact algebraic result by exploiting advances in solution algorithms and many-core computer architectures. Using state-of-the-art high performance dense linear algebra libraries associated with various leading edge parallel architectures (Intel KNLs, NVIDIA GPUs, and distributed-memory systems), ExaGeoStat raises the game for statistical applications from climate and environmental science. ExaGeoStat provides a reference evaluation of statistical parameters, with which to assess the validity of the various approaches based on approximation. The framework takes a first step in the merger of large-scale data analytics and extreme computing for geospatial statistical applications, to be followed by additional complexity reducing improvements from the solver side that can be implemented under the same interface. Thus, a single uncompromised statistical model can ultimately be executed in a wide variety of emerging exascale environments.

cs.DC↗