Some solutions to the multivariate Behrens–Fisher problem for dissimilarity‐based analyses

The essence of the generalised multivariate Behrens–Fisher problem (BFP) is how to test the null hypothesis of equality of mean vectors for two or more populations when their dispersion matrices differ. Solutions to the BFP usually assume variables are multivariate normal and do not handle high‐dimensional data. In ecology, species' count data are often high‐dimensional, non‐normal and heterogeneous. Also, interest lies in analysing compositional dissimilarities among whole communities in non‐Euclidean (semi‐metric or non‐metric) multivariate space. Hence, dissimilarity‐based tests by permutation (e.g., PERMANOVA, ANOSIM) are used to detect differences among groups of multivariate samples. Such tests are not robust, however, to heterogeneity of dispersions in the space of the chosen dissimilarity measure, most conspicuously for unbalanced designs. Here, we propose a modification to the PERMANOVA test statistic, coupled with either permutation or bootstrap resampling methods, as a solution to the BFP for dissimilarity‐based tests. Empirical simulations demonstrate that the type I error remains close to nominal significance levels under classical scenarios known to cause problems for the un‐modified test. Furthermore, the permutation approach is found to be more powerful than the (more conservative) bootstrap for detecting changes in community structure for real ecological datasets. The utility of the approach is shown through analysis of 809 species of benthic soft‐sediment invertebrates from 101 sites in five areas spanning 1960 km along the Norwegian continental shelf, based on the Jaccard dissimilarity measure.     

Projectors and linear estimation in general linear models

The paper gives a self-contained account of minimum disper­sion linear unbiased estimation of the expectation vector in a linear model with the dispersion matrix belonging to some, rather arbitrary, set of nonnegative definite matrices. The approach to linear estimation in general linear models recommended here is a direct generalization of some ideas and results presented by Rao (1973, 19 74) for the case of a general Gauss-Markov model

A new insight into the nature of some estimation problems originaly arising in the context of a general Gauss-Markov model as well as the correspondence of results known in the literature to those obtained in the present paper for general linear models are also given. As preliminary results the theory of projectors defined by Rao (1973) is extended.     

A comparison of two exact confidence regions for partially nonlinear regression models

Exact confidence regions for all the parameters in nonlinear regression models can be obtained by comparing the lengths of projections of the error vector into orthogonal subspaces of the sample space. In certain partially nonlinear models an alternative exact region is obtained by replacing the linear parameters by their conditional estimates in the projection matrices. An ellipsoidal approximation to the alternative region is obtained in terms of the tangent-plane coordinates, similar to one previously obtained for the more usual region. This ellipsoid can be converted to an approximate region for the original parameters and can be used to compare the two types of exact confidence regions.     

Reliability Evaluation of a Multi-state Network with Multiple Sinks under Individual Accuracy Rate Constraint

From the viewpoint of service level agreements (SLAs), Internet service providers and customers are gradually focusing on transmission accuracy. The Internet service provider should provide the specific bandwidth and individual accuracy rate requirement by their SLAs to each customer. This paper mainly evaluates the system reliability that a stochastic computer network can fulfill all requirements at all sinks. An efficient algorithm is proposed to generate the lower boundary points, minimal capacity vectors satisfying the demand and accuracy rate requirement for all sinks. The system reliability can be computed in terms of such points by applying recursive sum of disjoint products.     

Functional analysis of variance for Hilbert-valued multivariate fixed effect models

This paper presents new results on functional analysis of variance for fixed effect models with correlated Hilbert-valued Gaussian error components. The geometry of the reproducing kernel Hilbert space of the error term is considered in the computation of the total sum of squares, the residual sum of squares, and the sum of squares due to the regression. Under suitable linear transformation of the correlated functional data, the distributional characteristics of these statistics, their moment generating and characteristic functions, are derived. Fixed effect linear hypothesis testing is finally formulated in the Hilbert-valued multivariate Gaussian context considered.     

Exact posterior distributions and model selection criteria for multiple change-point detection problems

In segmentation problems, inference on change-point position and model selection are two difficult issues due to the discrete nature of change-points. In a Bayesian context, we derive exact, explicit and tractable formulae for the posterior distribution of variables such as the number of change-points or their positions. We also demonstrate that several classical Bayesian model selection criteria can be computed exactly. All these results are based on an efficient strategy to explore the whole segmentation space, which is very large. We illustrate our methodology on both simulated data and a comparative genomic hybridization profile.     

Quasi Bayesian likelihood

If the form of the distribution of data is unknown, the Bayesian method fails in the parametric inference because there is no posterior distribution of the parameter. In this paper, a theoretical framework of Bayesian likelihood is introduced via the Hilbert space method, which is free of the distributions of data and the parameter. The posterior distribution and posterior score function based on given inner products are defined and, consequently, the quasi posterior distribution and quasi posterior score function are derived, respectively, as the projections of the posterior distribution and posterior score function onto the space spanned by given estimating functions. In the space spanned by data, particularly, an explicit representation for the quasi posterior score function is obtained, which can be derived as a projection of the true posterior score function onto this space. The methods of constructing conservative quasi posterior score and quasi posterior log-likelihood are proposed. Some examples are given to illustrate the theoretical results. As an application, the quasi posterior distribution functions are used to select variables for generalized linear models. It is proved that, for linear models, the variable selections via quasi posterior distribution functions are equivalent to the variable selections via the penalized residual sum of squares or regression sum of squares.     

Predictive information and distance between past and future of a time series

A characterization for the nullity of the cosine angle between two subspaces of a Hilbert space is established. Given a time series x, we use this characterization in order to investigate the relationship between the notions of predictor space and distance between the information contained in the past and in the future of x. In particular, we prove that the predictor space of x coincides with the zero vector space {0} if and only if this distance achieves its maximum value.     

Stochastic Liénard Equations with Random Switching and Two-time Scales

This article is devoted to the study of stochastic Liénard equations with random switching. The motivation of our study stems from modeling of complex systems in which both continuous dynamics and discrete events are present. The continuous component is a solution of a stochastic Liénard equation and the discrete component is a Markov chain with a finite state space that is large. A distinct feature is that the processes under consideration are time inhomogeneous. Based on the idea of nearly decomposability and aggregation, the state space of the switching process can be viewed as “nearly decomposable” into l subspaces that are connected with weak interactions among the subspaces. Using the idea of aggregation, we lump the states in each subspace into a single state. Considering the pair of process (continuous state, discrete state), under suitable conditions, we derive a weak convergence result by means of martingale problem formulation. The significance of the limit process is that it is substantially simpler than that of the original system. Thus, it can be used in the approximation and computation work to reduce the computational complexity.     

Gauss–Christoffel quadrature for inverse regression: applications to computer experiments

Sufficient dimension reduction (SDR) provides a framework for reducing the predictor space dimension in statistical regression problems. We consider SDR in the context of dimension reduction for deterministic functions of several variables such as those arising in computer experiments. In this context, SDR can reveal low-dimensional ridge structure in functions. Two algorithms for SDR—sliced inverse regression (SIR) and sliced average variance estimation (SAVE)—approximate matrices of integrals using a sliced mapping of the response. We interpret this sliced approach as a Riemann sum approximation of the particular integrals arising in each algorithm. We employ the well-known tools from numerical analysis—namely, multivariate numerical integration and orthogonal polynomials—to produce new algorithms that improve upon the Riemann sum-based numerical integration in SIR and SAVE. We call the new algorithms Lanczos–Stieltjes inverse regression (LSIR) and Lanczos–Stieltjes average variance estimation (LSAVE) due to their connection with Stieltjes’ method—and Lanczos’ related discretization—for generating a sequence of polynomials that are orthogonal with respect to a given measure. We show that this approach approximates the desired integrals, and we study the behavior of LSIR and LSAVE with two numerical examples. The quadrature-based LSIR and LSAVE eliminate the first-order algebraic convergence rate bottleneck resulting from the Riemann sum approximation, thus enabling high-order numerical approximations of the integrals when appropriate. Moreover, LSIR and LSAVE perform as well as the best-case SIR and SAVE implementations (e.g., adaptive partitioning of the response space) when low-order numerical integration methods (e.g., simple Monte Carlo) are used.

     

Exhaustive k-nearest-neighbour subspace clustering

Cluster analysis is an important technique of explorative data mining. It refers to a collection of statistical methods for learning the structure of data by solely exploring pairwise distances or similarities. Often meaningful structures are not detectable in these high-dimensional feature spaces. Relevant features can be obfuscated by noise from irrelevant measurements. These observations led to the design of subspace clustering algorithms, which can identify clusters that originate from different subsets of features. Hunting for clusters in arbitrary subspaces is intractable due to the infinite search space spanned by all feature combinations. In this work, we present a subspace clustering algorithm that can be applied for exhaustively screening all feature combinations of small- or medium-sized datasets (approximately 30 features). Based on a robustness analysis via subsampling we are able to identify a set of stable candidate subspace cluster solutions.     

Bayes Hilbert Spaces

A Bayes linear space is a linear space of equivalence classes of proportional σ‐finite measures, including probability measures. Measures are identified with their density functions. Addition is given by Bayes' rule and substraction by Radon–Nikodym derivatives. The present contribution shows the subspace of square‐log‐integrable densities to be a Hilbert space, which can include probability and infinite measures, measures on the whole real line or discrete measures. It extends the ideas from the Hilbert space of densities on a finite support towards Hilbert spaces on general measure spaces. It is also a generalisation of the Euclidean structure of the simplex, the sample space of random compositions. In this framework, basic notions of mathematical statistics get a simple algebraic interpretation. A key tool is the centred‐log‐ratio transformation, a generalization of that used in compositional data analysis, which maps the Hilbert space of measures into a subspace of square‐integrable functions. As a consequence of this structure, distances between densities, orthonormal bases, and Fourier series representing measures become available. As an application, Fourier series of normal distributions and distances between them are derived, and an example related to grain size distributions is presented. The geometry of the sample space of random compositions, known as Aitchison geometry of the simplex, is obtained as a particular case of the Hilbert space when the measures have discrete and finite support.     

Consistent inference for biased sub-model of high-dimensional partially linear model

In this paper, we study a working sub-model of partially linear model determined by variable selection. Such a sub-model is more feasible and practical in application, but usually biased. As a result, the common parameter estimators are inconsistent and the corresponding confidence regions are invalid. To deal with the problems relating to the model bias, a nonparametric adjustment procedure is provided to construct a partially unbiased sub-model. It is proved that both the adjusted restricted-model estimator and the adjusted preliminary test estimator are partially consistent, which means when the samples drop into some given subspaces, the estimators are consistent. Luckily, such subspaces are large enough in a certain sense and thus such a partial consistency is close to global consistency. Furthermore, we build a valid confidence region for parameters in the sub-model by the corresponding empirical likelihood.     

Signatures of series and parallel systems consisting of non disjoint modules

System signature is a useful tool to analyze coherent systems. In reliability theory, a large number of systems is actually the composition of disjoint subsystems (modules). However, in real life, there are situations in which the subsystems have common components. In this article, we consider the problem of obtaining signatures as well as minimal signatures of series and parallel systems that are composed of subsystems sharing a component. That is, these subsystems are no longer disjoint. Computational results are also presented to illustrate our findings.     

Non-minimaxity of linear combinations of restricted location estimators and related problems

The estimation of a linear combination of several restricted location parameters is addressed from a decision-theoretic point of view. Although the corresponding linear combination of the unbiased estimators is minimax under the restricted problem, it has a drawback of taking values outside the restricted parameter space. Thus, it is reasonable to use the linear combination of the restricted estimators such as maximum likelihood or truncated estimators. In this paper, a necessary and sufficient condition for such restricted estimators to be minimax is derived, and it is shown that the restricted estimators are not minimax when the number of the location parameters is large. The condition for minimaxity is examined for some specific distributions. Finally, similar problems of estimating the product and sum of the restricted scale parameters are studied, and it is shown that analogous non-dominance properties appear when the number of the scale parameters is large.     

Robustness of designed experiments against missing data

This paper investigates the robustness of designed experiments for estimating linear functions of a subset of parameters in a general linear model against the loss of any t( ≥1) observations. Necessary and sufficient conditions for robustness of a design under a homoscedastic model are derived. It is shown that a design robust under a homoscedastic model is also robust under a general heteroscedastic model with correlated observations. As a particular case, necessary and sufficient conditions are obtained for the robustness of block designs against the loss of data. Simple sufficient conditions are also provided for the binary block designs to be robust against the loss of data. Some classes of designs, robust up to three missing observations, are identified. A-efficiency of the residual design is evaluated for certain block designs for several patterns of two missing observations. The efficiency of the residual design has also been worked out when all the observations in any two blocks, not necessarily disjoint, are lost. The lower bound to A-efficiency has also been obtained for the loss of t observations. Finally, a general expression is obtained for the efficiency of the residual design when all the observations of m ( ≥1) disjoint blocks are lost.     

Current and past roles of the statistician in space applications

The paper contains a brief and informal history of space science topics which led to statistical consulting. The past and current roles of the statistician in space science is discussed and this special issue is prefaced by describing generally the contents of the issue. Several statistical problems associated with the space scientists’ efforts to develop instrumentation and space vehicles for both manned and unmanned space missions are discussed as well as those problems associated with developing a space laboratory. The current tasks of developing remote sensors to make earth observations for monitoring earth resources and the associated statistical problems are also discussed. These latter activities are those that have motivated this special issue.     

Exact filtering for partially observed continuous time models

Summary.  The forward–backward algorithm is an exact filtering algorithm which can efficiently calculate likelihoods, and which can be used to simulate from posterior distributions. Using a simple result which relates gamma random variables with different rates, we show how the forward–backward algorithm can be used to calculate the distribution of a sum of gamma random variables, and to simulate from their joint distribution given their sum. One application is to calculating the density of the time of a specific event in a Markov process, as this time is the sum of exponentially distributed interevent times. This enables us to apply the forward–backward algorithm to a range of new problems. We demonstrate our method on three problems: calculating likelihoods and simulating allele frequencies under a non-neutral population genetic model, analysing a stochastic epidemic model and simulating speciation times in phylogenetics.     

Wavelet kernel penalized estimation for non-equispaced design regression

The paper considers regression problems with univariate design points. The design points are irregular and no assumptions on their distribution are imposed. The regression function is retrieved by a wavelet based reproducing kernel Hilbert space (RKHS) technique with the penalty equal to the sum of blockwise RKHS norms. In order to simplify numerical optimization, the problem is replaced by an equivalent quadratic minimization problem with an additional penalty term. The computational algorithm is described in detail and is implemented with both the sets of simulated and real data. Comparison with existing methods showed that the technique suggested in the paper does not oversmooth the function and is superior in terms of the mean squared error. It is also demonstrated that under additional assumptions on design points the method achieves asymptotic optimality in a wide range of Besov spaces.