Modeling the coverage probability of a confidence interval on the among-group variance component in the unbalanced random one-way model

The coverage probability of an approximate confidence interval on the among-group variance component, σ _α ² , in a one-way random model is modeled using generalized linear models techniques. The purpose of the proposed modeling is to derive an empirical relationship between the coverage probability on one hand, andk, n., ?, and the model’s variance components on the other hand, wherek is the number of groups, n. is the total number of observations, and ? is a measure of imbalance for the design used. The latter quantities serve as control variables in the derived model, and the coverage probability is treated as the response variable. Contour plots generated from this model can be easily used to depict the effects of the control variables on the coverage probability. In particular, the plots are utilized to compare four methods for constructing approximate confidence intervals on σ _α ² . Additional advantages of the derived model include prediction of the coverage probability for a given method using only specified values of the control variables, and the determination of operating conditions that result in improved coverage probability within the region of interest.     

The credit risk+ model with general sector correlations

We consider an enhancement of the credit risk⁺ model to incorporate correlations between sectors. We model the sector default rates as linear combinations of a common set of independent variables that represent macro-economic variables or risk factors. We also derive the formula for exact VaR contributions at the obligor level.     

A bipolar model of assertability and belief

Valuation pairs are introduced as a bipolar model of the assertability of propositions. These correspond to a pair of dual valuation functions, respectively, representing the strong property of definite assertability and the dual weaker property of acceptable assertability. In the case where there is uncertainty about the correct valuation pair for a language then a probability distribution is defined on possible valuation pairs. This results in two measures, μ⁺ giving the probability that a sentence is definitely assertable, and μ⁻ giving the probability that a sentence is acceptable to assert. It is shown that μ⁺ and μ⁻ can be determined directly from a two dimensional mass function m defined on pairs of sets of propositional variables. Certain natural properties of μ⁺ and μ⁻ are easily expressed in terms of m, and in particular we introduce certain consonance or nestedness assumptions. These capture qualitative information in the form of assertability orderings for both the propositional variables and the negated propositional variables. On the basis of these consonance assumptions we show that label semantics, intuitionistic fuzzy logic and max-min fuzzy logic can all be viewed as special cases of this bipolar model. We also show that bipolar belief measures can be interpreted within an interval-set model.     

A higher order finite element including transverse normal strain for linear elastic composite plates with general lamination configurations

This paper describes a higher-order global-local theory for thermal/mechanical response of moderately thick laminated composites with general lamination configurations. In-plane displacement fields are constructed by superimposing the third-order local displacement field to the global cubic displacement field. To eliminate layer-dependent variables, interlaminar shear stress compatibility conditions have been employed, so that the number of variables involved in the proposed model is independent of the number of layers of laminates. Imposing shear stress free condition at the top and the bottom surfaces, derivatives of transverse displacement are eliminated from the displacement field, so that C⁰ interpolation functions are only required for the finite element implementation. To assess the proposed model, the quadratic six-node C⁰ triangular element is employed for the interpolation of all the displacement parameters defined at each nodal point on the composite plate. Comparing to various existing laminated plate models, it is found that simple C⁰ finite elements with non-zero normal strain could produce more accurate displacement and stresses for thick multilayer composite plates subjected to thermal and mechanical loads. Finally, it is remarked that the proposed model is quite robust, such that the finite element results are not sensitive to the mesh configuration and can rapidly converge to 3-D elasticity solutions using regular or irregular meshes.     

Binary clustering with missing data

A clustering method is presented for analysing multivariate binary data with missing values. When not all values are observed, Govaert³ has studied the relations between clustering methods and statistical models. The author has shown how the identification of a mixture of Bernoulli distributions with the same parameter for all clusters and for all variables corresponds to a clustering criterion which uses L₁ distance characterizing the MNDBIN method (Marchetti⁸). He first generalized this model by selecting parameters which can depend on variables and finally by selecting parameters which can depend both on variables and on clusters. We use the previous models to derive a clustering method adapted to missing data. This method optimizes a criterion by a standard iterative partitioning algorithm which removes the necessity either to ignore objects or to substitute the missing data. We study several versions of this algorithm and, finally, a brief account is given of the application of this method to some simulated data.     

Toral algebraic sets and function theory on polydisks

A toral algebraic set A is an algebraic set in ℂ ⁿ whose intersection with T ⁿ is sufficiently large to determine the holomorphic functions on A. We develop the theory of these sets, and give a number of applications to function theory in several variables and operator theoretic model theory. In particular, we show that the uniqueness set for an extremal Pick problem on the bidisk is a toral algebraic set, that rational inner functions have zero sets whose irreducible components are not toral, and that the model theory for a commuting pair of contractions with finite defect lives naturally on a toral algebraic set.     

Experimental study on the efficiency and accuracy of a chance-constrained programming algorithm

The CHAPS algorithm (CHAPS = Chance-Constrained Programming System) has proved to be an efficient and accurate method for solving linear optimization problems which have several random variables distributed normally and independently of each other. The CHAPS algorithm is based on the separation, linearization and iterative adjusting of linearization of chance-constrained deterministic equivalents by using the simplex method.According to test results the solution time of the algorithm is directly proportional to the second power of the number of constraints of a linearized model corresponding to the chance-constrained model. The positive result is partly due to the fact that the linearized model is very sparse. The algorithm requires six to eight CHAPS iteration runs in order to achieve sufficient accuracy in practice (10^?5 –10^?6). The algorithm converges linearly and its asymptotic error constant is $\text{1}\text{4}$.     

Comparing the Value at Risk Performance of the CreditRisk + and its Enhancement: A Large Deviations Approach

The standard CreditRisk^?+? (CR^?+?) is a well-known default-mode credit risk model. An extension to the CR^?+? that introduces correlation through a two-stage hierarchy of randomness has been discussed by Deshpande and Iyer (Central Eur J Oper Res 17(2):219–228, 2009) and more recently by Sowers (2010). It is termed the 2-stage CreditRisk^?+? (2-CR^?+?) in the former. Unlike the standard CR^?+?, the 2-CR^?+? model is formulated to allow correlation between sectoral default rates through dependence on a common set of macroeconomic variables. Furthermore the default rates for a 2-CR^?+? are distributed according to a general univariate distribution which is in stark contrast to the uniformly Gamma distributed sectoral default rates in the CR^?+?. We would then like to understand the behaviour of these two models with regards to their computed Value at Risk (VaR) as the number of sectors and macroeconomic variables approaches infinity. In particular we would like to ask whether the 2-CR^?+? produces higher VaR than the CR^?+? and if so then for which type of credit portfolio. Utilizing the theory of Large deviations, we provide a methodology for comparing the Value at risk performance of these two competing models by computing certain associated rare event probabilities. In particular we show that the 2-Stage CR^?+? definitely produces higher VaR than the CR^?+? for a particular class of a credit portfolio which we term as a “balanced” credit portfolio. We support this statistical risk analysis through numerical examples.     

The economy and loss given default: evidence from two UK retail lending data sets

Loss given default (LGD) models predict losses as a proportion of the outstanding loan, in the event a debtor goes into default. The literature on corporate sector LGD models suggests LGD is correlated to the economy and so changes in the economy could translate into different predictions of losses. In this work, the role of macroeconomic variables in loan-level retail LGD models is examined by testing the inclusion of macroeconomic variables in two different retail LGD models: a two-stage model for a residential mortgage loans data set and an ordinary least squares model for an unsecured personal loans data set. To improve loan-level predictions of LGD, indicators relating to the macroeconomy are considered with mixed results: the selected macroeconomic variable seemed able to improve the predictive performance of mortgage loan LGD estimates, but not for personal loan LGD. For mortgage loan LGD, interest rate was most beneficial but only predicted better during downturn periods, underestimating LGD during non-downturn periods. For personal loan LGD, only net lending growth is statistically significant but including this variable did not bring any improvement to R².     

On a characterization of orthogonality with respect to particular sequences of random variables in <Emphasis Type="Italic">L</Emphasis><Superscript>2</Superscript>

This note deals with the orthogonality between sequences of random variables. The main idea of the note is to apply the results on equidistant systems of points in a Hilbert space to the case of the space L ²(Ω, F, ℙ) of real square integrable random variables. The main result gives a necessary and sufficient condition for a particular sequence of random variables (elements of which are taken from sets of equidistant elements of L ²(Ω, F, ℙ) to be orthogonal to some other sequence in L ²(Ω, F, ℙ). The result obtained is interesting from the point of view of the time series analysis, since it can be applied to a class of sequences random variables that exhibit a monotonically increasing variance. An application to ergodic theorem is also provided.     

Asymptotic distribution of the OLS estimator for a purely autoregressive spatial model

We derive the asymptotics of the OLS estimator for a purely autoregressive spatial model. Only low-level conditions are used. As the sample size increases, the spatial matrix is assumed to approach a square-integrable function on the square (0,1)². The asymptotic distribution is a ratio of two infinite linear combinations of χ² variables. The formula involves eigenvalues of an integral operator associated with the function approached by the spatial matrices. Under the conditions imposed identification conditions for the maximum likelihood method and method of moments fail. A corrective two-step procedure using the OLS estimator is proposed.     

A linear regression model for imprecise response

A linear regression model with imprecise response and p real explanatory variables is analyzed. The imprecision of the response variable is functionally described by means of certain kinds of fuzzy sets, the LR fuzzy sets. The LR fuzzy random variables are introduced to model usual random experiments when the characteristic observed on each result can be described with fuzzy numbers of a particular class, determined by 3 random values: the center, the left spread and the right spread. In fact, these constitute a natural generalization of the interval data. To deal with the estimation problem the space of the LR fuzzy numbers is proved to be isometric to a closed and convex cone of R³ with respect to a generalization of the most used metric for LR fuzzy numbers. The expression of the estimators in terms of moments is established, their limit distribution and asymptotic properties are analyzed and applied to the determination of confidence regions and hypothesis testing procedures. The results are illustrated by means of some case-studies.     

Exactly solvable two-dimensional complex model with a real spectrum

Using supersymmetric intertwining relations of the second order in derivatives, we construct a two-dimensional quantum model with a complex potential for which all energy levels and the corresponding wave functions are obtained analytically. This model does not admit separation of variables and can be considered a complexified version of the generalized two-dimensional Morse model with an additional sinh ⁻² term. We prove that the energy spectrum of the model is purely real. To our knowledge, this is a rather rare example of a nontrivial exactly solvable model in two dimensions. We explicitly find the symmetry operator, describe the biorthogonal basis, and demonstrate the pseudo-Hermiticity of the Hamiltonian of the model. The obtained wave functions are simultaneously eigenfunctions of the symmetry operator. Dedicated to the 80th birthday of Yuri Victorovich Novozhilov __________ Translated from Teoreticheskaya i Matematicheskaya Fizika, Vol. 148, No. 1, pp. 102–111, July, 2006.     

On random orderings of variables for parity ordered binary decision diagrams

Ordered binary decision diagrams (OBDDs) are a model for representing Boolean functions. There is also a more powerful variant called parity OBDDs. The size of the representation of a given function depends in both these models on the chosen ordering of the variables. It is known that there are functions such that almost all orderings of their variables yield an OBDD of polynomial size, but there are also some exceptional orderings, for which the size is exponential. We prove that for parity OBDDs, the size for a random ordering and the size for the worst ordering are polynomially related. More exactly, for every ϵ>0 there is a number c>0 such that the following holds. If a Boolean function f of n variables is such that a random ordering of the variables yields a parity OBDD for f of size at most s with probability at least ϵ, where s≥n, then every ordering of the variables yields a parity OBDD for f of size at most s^c. © 2000 John Wiley & Sons, Inc. Random Struct. Alg., 16: 233–239, 2000     

On χ2-test of goodness-of-fit for a class of discrete multivariate models

A model for an n-dimensional random vector which is the sum of v i.i.d. random variables with an exponential family distribution is considered. It is shown that the exact test can be approximated by a χ²-test as v → ∞. As an example a goodness-of-fit test for Poisson-distribution based on observed frequencies is derived.     

Asymptotic expansions for anisotropic heat kernels

We obtain the asymptotic expansion of the solutions of some anisotropic heat equations when the initial data belong to polynomially weighted L ^p -spaces. We mainly address two model examples. In the first one, the diffusivity is of order two in some variables but higher in the other ones. In the second one, we consider the heat equation on the Heisenberg group.