Latent class analysis variable selection

We propose a method for selecting variables in latent class analysis, which is the most common model-based clustering method for discrete data. The method assesses a variable’s usefulness for clustering by comparing two models, given the clustering variables already selected. In one model the variable contributes information about cluster allocation beyond that contained in the already selected variables, and in the other model it does not. A headlong search algorithm is used to explore the model space and select clustering variables. In simulated datasets we found that the method selected the correct clustering variables, and also led to improvements in classification performance and in accuracy of the choice of the number of classes. In two real datasets, our method discovered the same group structure with fewer variables. In a dataset from the International HapMap Project consisting of 639 single nucleotide polymorphisms (SNPs) from 210 members of different groups, our method discovered the same group structure with a much smaller number of SNPs.     

On the exploration of linear latent effect for multivariate modeling

A data analysis method is proposed to derive a latent structure matrix from a sample covariance matrix. The matrix can be used to explore the linear latent effect between two sets of observed variables. Procedures with which to estimate a set of dependent variables from a set of explanatory variables by using latent structure matrix are also proposed. The proposed method can assist the researchers in improving the effectiveness of the SEM models by exploring the latent structure between two sets of variables. In addition, a structure residual matrix can also be derived as a by-product of the proposed method, with which researchers can conduct experimental procedures for variables combinations and selections to build various models for hypotheses testing. These capabilities of data analysis method can improve the effectiveness of traditional SEM methods in data property characterization and models hypotheses testing. Case studies are provided to demonstrate the procedure of deriving latent structure matrix step by step, and the latent structure estimation results are quite close to the results of PLS regression. A structure coefficient index is suggested to explore the relationships among various combinations of variables and their effects on the variance of the latent structure.     

Learning the Structure of Mixed Graphical Models

We consider the problem of learning the structure of a pairwise graphical model over continuous and discrete variables. We present a new pairwise model for graphical models with both continuous and discrete variables that is amenable to structure learning. In previous work, authors have considered structure learning of Gaussian graphical models and structure learning of discrete models. Our approach is a natural generalization of these two lines of work to the mixed case. The penalization scheme involves a novel symmetric use of the group-lasso norm and follows naturally from a particular parameterization of the model. Supplementary materials for this article are available online.     

Multihazard Design: Structural Optimization Approach

The objective of multihazard structural engineering is to develop methodologies for achieving designs that are safe and cost-effective under multiple hazards. Optimization is a natural tool for achieving such designs. In general, its aim is to determine a vector of design variables subjected to a given set of constraints, such that an objective function of those variables is minimized. In the particular case of structural design, the design variables may be member sizes; the constraints pertain to structural strength and serviceability (e.g., keeping the load-induced stresses and deflections below specified thresholds); and the objective function is the structure cost or weight. In a multihazard context, the design variables are subjected to the constraints imposed by all the hazards to which the structure is exposed. In this paper, we formulate the multihazard structural design problem in nonlinear programming terms and present a simple illustrative example involving four design variables and two hazards: earthquake and strong winds. Results of our numerical experiments show that interior-point methods are significantly more efficient than classical optimization methods in solving the nonlinear programming problem associated with our illustrative example.     

Identification of graphical models for nonignorable nonresponse of binary outcomes in longitudinal studies

In this paper, we use directed acyclic graphs (DAGs) with temporal structure to describe models of nonignorable nonresponse mechanisms for binary outcomes in longitudinal studies, and we discuss identification of these models under an assumption that the sequence of variables has the first-order Markov dependence, that is, the future variables are independent of the past variables conditional on the present variables. We give a stepwise approach for checking identifiability of DAG models. For an unidentifiable model, we propose adding completely observed variables such that this model becomes identifiable.     

�Ǿ���Գ��Զ���Ϊ3/n(n>=3)����������Ľṹ������ѽ�

We study the random variables of radial asymmetry based on copulas. We research on the structure of random variables which radial asymmetry degree isand get the exact best-possible bounds of random variables which radial asymmetry degree is equal to. Then we expand to general case. We propose an essential condition of radial asymmetry degree is and study the structure of copula. We provide a broad bounds of copula that the radial asymmetry degree is .     

A heuristic algorithm for a chance constrained stochastic program

A chance constrained stochastic program is considered that arises from an application to college enrollments and in which the objective function is the expectation of a linear function of the random variables. When these random variables are independent and normally distributed with mean and variance that are linear in the decision variables, the deterministic equivalent of the problem is a nonconvex nonlinear knapsack problem. The optimal solution to this problem is characterized and a greedy-type heuristic algorithm that exploits this structure is employed. Computational results show that the algorithm performs well, especially when the normal random variables are approximations of binomial random variables.     

Supervised clustering of variables

In predictive modelling, highly correlated predictors lead to unstable models that are often difficult to interpret. The selection of features, or the use of latent components that reduce the complexity among correlated observed variables, are common strategies. Our objective with the new procedure that we advocate here is to achieve both purposes: to highlight the group structure among the variables and to identify the most relevant groups of variables for prediction. The proposed procedure is an iterative adaptation of a method developed for the clustering of variables around latent variables (CLV). Modification of the standard CLV algorithm leads to a supervised procedure, in the sense that the variable to be predicted plays an active role in the clustering. The latent variables associated with the groups of variables, selected for their “proximity” to the variable to be predicted and their “internal homogeneity”, are progressively added in a predictive model. The features of the methodology are illustrated based on a simulation study and a real-world application.     

Discovering Sparse Covariance Structures With the Isomap

Regularization of covariance matrices in high dimensions usually either is based on a known ordering of variables or ignores the ordering entirely. This article proposes a method for discovering meaningful orderings of variables based on their correlations using the Isomap, a nonlinear dimension reduction technique designed for manifold embeddings. These orderings are then used to construct a sparse covariance estimator, which is block-diagonal and/or banded. Finding an ordering to which banding can be applied is desirable because banded estimators have been shown to be consistent in high dimensions. We show that in situations where the variables do have such a structure, the Isomap does very well at discovering it, and the resulting regularized estimator performs better for covariance estimation than other regularization methods that ignore variable order, such as thresholding. We also propose a bootstrap approach to constructing the neighborhood graph used by the Isomap, and show it leads to better estimation. We illustrate our method on data on protein consumption, where the variables (food types) have a structure but it cannot be easily described a priori, and on a gene expression dataset. Supplementary materials are available online.     

Stress testing network reconstruction via graphical causal model

An resilience optimal evaluation of financial portfolios implies having plausible hypotheses about the multiple interconnections between the macroeconomic variables and the risk parameters. In this article, we propose a graphical model for the reconstruction of the causal structure that links the multiple macroeconomic variables and the assessed risk parameters, it is this structure that we call stress testing network. In this model, the relationships between the macroeconomic variables and the risk parameter define a “relational graph” among their time‐series, where related time‐series are connected by an edge. Our proposal is based on the temporal causal models, but unlike, we incorporate specific conditions in the structure which correspond to intrinsic characteristics this type of networks. Using the proposed model and given the high‐dimensional nature of the problem, we used regularization methods to efficiently detect causality in the time‐series and reconstruct the underlying causal structure. In addition, we illustrate the use of model in credit risk data of a portfolio. Finally, we discuss its uses and practical benefits in stress testing.     

On the distribution of the (un)bounded sum of random variables

We propose a general treatment of random variables aggregation accounting for the dependence among variables and bounded or unbounded support of their sum. The approach is based on the extension to the concept of convolution to dependent variables, involving copula functions. We show that some classes of copula functions (such as Marshall-Olkin and elliptical) cannot be used to represent the dependence structure of two variables whose sum is bounded, while Archimedean copulas can be applied only if the generator becomes linear beyond some point. As for the application, we study the problem of capital allocation between risks when the sum of losses is bounded.     

A simple Tabu Search method to solve the mixed-integer linear bilevel programming problem

Multilevel programming is characterized as mathematical programming to solve decentralized planning problems. The models partition control over decision variables among ordered levels within a hierarchical planning structure of which the linear bilevel form is a special case of a multilevel programming problem. In a system with such a hierarchical structure, the high-level decision making situations generally require inclusion of zero-one variables representing ‘yes-no’ decisions. We provide a mixed-integer linear bilevel programming formulation in which zero-one decision variables are controlled by a high-level decision maker and real-value decision variables are controlled by a low-level decision maker. An algorithm based on the short term memory component of Tabu Search, called Simple Tabu Search, is developed to solve the problem, and two supplementary procedures are proposed that provide variations of the algorithm. Computational results disclose that our approach is effective in terms of both solution quality and efficiency.     

利率期限结构与宏观经济变量的相互关系研究

本文研究了利率期限结构与宏观经济变量之间的相互关系。运用利率期限结构与宏观经济变量的无套利模型,对向量自回归模型进行了扩展,将其引入到状态空间模型框架中,基于卡尔曼滤波并结合EM算法对模型参数进行了有效估计,结合实际数据对利率期限结构与宏观经济变量的相互影响关系进行了实证研究。结果表明:利率期限结构与宏观经济变量的双向影响关系显著;宏观经济变量对利率期限结构具有一定的解释力;研究利率期限结构时,宏观经济变量的影响作用不能忽略。     

A simple,all primal branch and bound approach to pure and mixed integer binary programs

Many branch and bound procedures for integer programming employ linear programming to obtain bound information. Nodes in the tree structure are defined by explicitly changing bounds on certain variables and/or adding one or more constraints to the parent LP; thus, primal feasibility is destroyed. The design and analysis of the resulting tree structure requires that basis information be stored for each node and that feasibility restoring pivots be used to obtain the node bound. In turn, this may require the introduction of artificial variables and/or dual simplex pivots.This paper describes a simple procedure for branch and bound that does not destroy primal feasibility. Moreover, the information required to be stored to define the node problems is minimal.     

On theories of random variables

We study theories of spaces of random variables: first, we consider random variables with values in the interval [0, 1], then with values in an arbitrary metric structure, generalising Keisler’s randomisation of classical structures. We prove preservation and non-preservation results for model theoretic properties under this construction:

1. The randomisation of a stable structure is stable.
2. The randomisation of a simple unstable structure is not simple.

We also prove that in the randomised structure, every type is a Lascar type.     

Extracting informative variables in the validation of two-group causal relationship

The validation of causal relationship between two groups of multivariate time series data often requires the precedence knowledge of all variables. However, in practice one finds that some variables may be negligible in describing the underlying causal structure. In this article we provide an explicit definition of “non-informative variables” in a two-group causal relationship and introduce various automatic computer-search algorithms that can be utilized to extract informative variables based on a hypothesis testing procedure. The result allows us to represent a simplified causal relationship by using minimum possible information on two groups of variables.     

On the structure of solutions of a class of hyperbolic systems with several spatial variables in the far field

An asymptotic expansion of the solution to the Cauchy problem for a class of hyperbolic weakly nonlinear systems with many spatial variables is constructed. A parabolic quasilinear equation describing the behavior of the solution at asymptotically large values of the independent variables is obtained. The pseudo-diffusion processes that depend on the relationship between the number of equations and the number of spatial variables are analyzed. The structure of the subspace in which there are pseudo-diffusion evolution processes of the solution in the far field is described.     

Time response of structure with interval and random parameters using a new hybrid uncertain analysis method

Practical structures often operate with some degree of uncertainties, and the uncertainties are often modelled as random parameters or interval parameters. For realistic predictions of the structures behaviour and performance, structure models should account for these uncertainties. This paper deals with time responses of engineering structures in the presence of random and/or interval uncertainties. Three uncertain structure models are introduced. The first one is random uncertain structure model with only random variables. The generalized polynomial chaos (PC) theory is applied to solve the random uncertain structure model. The second one is interval uncertain structure model with only interval variables. The Legendre metamodel (LM) method is presented to solve the interval uncertain structure model. The LM is based on Legendre polynomial expansion. The third one is hybrid uncertain structure model with both random and interval variables. The polynomial-chaos-Legendre-metamodel (PCLM) method is presented to solve the hybrid uncertain structure model. The PCLM is a combination of PC and LM. Three engineering examples are employed to demonstrate the effectiveness of the proposed methods. The uncertainties resulting from geometrical size, material properties or external loads are studied.     

Complete moment convergence of pairwise NQD random variables

It is known that the dependence structure of pairwise negative quadrant dependent (NQD) random variables is weaker than those of negatively associated random variables and negatively orthant dependent random variables. In this article, we investigate the moving average process which is based on the pairwise NQD random variables. The complete moment convergence and the integrability of the supremum are presented for this moving average process. The results imply complete convergence and the Marcinkiewicz–Zygmund-type strong law of large numbers for pairwise NQD sequences.     

The almost equivalence of pairwise and mutual independence and the duality with exchangeability

For a large collection of random variables in an ideal setting, pairwise independence is shown to be almost equivalent to mutual independence. An asymptotic interpretation of this fact shows the equivalence of asymptotic pairwise independence and asymptotic mutual independence for a triangular array (or a sequence) of random variables. Similar equivalence is also presented for uncorrelatedness and orthogonality as well as for the constancy of joint moment functions and exchangeability. General unification of multiplicative properties for random variables are obtained. The duality between independence and exchangeability is established through the random variables and sample functions in a process. Implications in other areas are also discussed, which include a justification for the use of mutually independent random variables derived from sequential draws where the underlying population only satisfies a version of weak dependence. Macroscopic stability of some mass phenomena in economics is also characterized via almost mutual independence. It is also pointed out that the unit interval can be used to index random variables in the ideal setting, provided that it is endowed together with some sample space a suitable larger measure structure. Received: 16 April 1997 / Revised version: 18 May 1998