Riesz inverse Gaussian distributions

The notion of generalized power of a positive definite symmetric matrix and a related notion of generalized Bessel function are used to introduce an extension of the class of matrix generalized inverse Gaussian distributions. The new distributions are shown to arise as conditional distributions of Peirce components of Riesz random matrices. Things are explained in the modern framework of symmetric cones and simple Euclidean Jordan algebra.     

Distribution of the length of the longest common subsequence of two multi-state biological sequences

The length of the longest common subsequence (LCS) among two biological sequences has been used as a measure of similarity, and the application of this statistic is of importance in genomic studies. Even for the simple case of two sequences of equal length and composed of binary elements with equal state probabilities, the exact distribution of the length of the LCS remains an open question. This problem is also known as an NP-hard problem in computer science. Apart from combinatorial analysis, using the finite Markov chain imbedding technique, we derive the exact distribution for the length of the LCS between two multi-state sequences of different lengths. Numerical results are provided to illustrate the theoretical results.     

Model Selection,Transformations and Variance Estimation in Nonlinear Regression

The results of analyzing experimental data using a parametric model may heavily depend on the chosen model for regression and variance functions, moreover also on a possibly underlying preliminary transformation of the variables. In this paper we propose and discuss a complex procedure which consists in a simultaneous selection of parametric regression and variance models from a relatively rich model class and of Box-Cox variable transformations by minimization of a cross-validation criterion. For this it is essential to introduce modifications of the standard cross-validation criterion adapted to each of the following objectives: 1. estimation of the unknown regression function, 2. prediction of future values of the response variable, 3. calibration or 4. estimation of some parameter with a certain meaning in the corresponding field of application. Our idea of a criterion oriented combination of procedures (which usually if applied, then in an independent or sequential way) is expected to lead to more accurate results. We show how the accuracy of the parameter estimators can be assessed by a “moment oriented bootstrap procedure", which is an essential modification of the “wild bootstrap” of Härdle and Mammen by use of more accurate variance estimates. This new procedure and its refinement by a bootstrap based pivot (“double bootstrap”) is also used for the construction of confidence, prediction and calibration intervals. Programs written in Splus which realize our strategy for nonlinear regression modelling and parameter estimation are described as well. The performance of the selected model is discussed, and the behaviour of the procedures is illustrated, e.g., by an application in radioimmunological assay.     

Estimating the distribution of one-dimensional discrete scan statistics viewed as extremes of 1-dependent stationary sequences

Discrete one-dimensional scan statistics can be viewed as extremes of 1-dependent stationary sequences. A result of Haiman [1999. First passage time for some stationary processes. Stochastic Process. Appl. 80, 231–248] provides approximations of the distributions of extremes of 1-dependent stationary sequences together with sharp bounds for the corresponding errors. We apply this result to scan statistics generated by Bernoulli r.v.'s and to the charge problem.     

A Comparison of Time-Varying Parameter and Multiprocess Mixture Models in the Case of Money-Supply Announcements

This study compares the performance of a recently proposed multiprocess mixture model and a random-walk time-varying parameter (TVP) model, using the interest rate–weekly money relationship for illustrative purposes. For the case of this relationship, which is subject to regime shifts and outliers, the mixture model performs well and the latter model performs poorly. This finding is of general interest, since investigators often adopt random-walk TVP models to accommodate potential regime shifts in regression relationships. The TVP estimation procedure is unlikely to find abrupt shifts, since the estimate of parameter variance is based on the entire data sample. In the face of rapid discontinuous shifts in the parameters, this variance estimate is unrepresentative of the variability during periods of abrupt shift or transient observations.     

Variance estimation in the central limit theorem for Markov chains

This article concerns the variance estimation in the central limit theorem for finite recurrent Markov chains. The associated variance is calculated in terms of the transition matrix of the Markov chain. We prove the equivalence of different matrix forms representing this variance. The maximum likelihood estimator for this variance is constructed and it is proved that it is strongly consistent and asymptotically normal. The main part of our analysis consists in presenting closed matrix forms for this new variance. Additionally, we prove the asymptotic equivalence between the empirical and the maximum likelihood estimation (MLE) for the stationary distribution.     

Perfect linear models and perfect parametric functions

General linear models with a common design matrix and with various structures of the variance–covariance matrix are considered. We say that a model is perfect for a linearly estimable parametric function, or the function is perfect in the model, if there exists the best linear unbiased estimator. All perfect models for a given function and all perfect functions in a given model are characterized.     

A note on replacement policy comparisons from NBUC lifetime of the unit

Age and block replacement policies are commonly used in order to reduce the number of in-service failures when the systems are functioning indefinitely. In reliability theory, the lifetime of a system can be modeled by means of the NBUC aging class that is characterized throughout comparisons of the residual lives in the sense of the icx order. The purpose of this paper is to establish stochastic comparisons between the age (block) replacement policy and a renewal process with no planned replacements when the lifetime of the unit is NBUC. Supported by Ministerio de Ciencia y Tecnología under grant BFM2000-0362     

Pitfalls in the Estimation of a Differenced Model

This article assesses the potential magnitude of the loss of estimation efficiency caused by the adoption of a differenced model when the disturbances of the original (levels) linear regression model follow either a stable (autoregressive) AR(1) process or a fixed start-up random-walk process (hence no filtering is necessary from the standpoint of estimation). The magnitude of the loss, which can be quite large, is found to be affected by both the form of the original model (homogeneous or nonhomogeneous) and the sign and magnitude of the autocorrelation coefficient of the AR(1) disturbance, as well as by the nature of the exogenous variable (smoothly trended or not).     

Likelihood-based inference for a class of multivariate diffusions with unobserved paths

This paper presents a Markov chain Monte Carlo algorithm for a class of multivariate diffusion models with unobserved paths. This class is of high practical interest as it includes most diffusion driven stochastic volatility models. The algorithm is based on a data augmentation scheme where the paths are treated as missing data. However, unless these paths are transformed so that the dominating measure is independent of any parameters, the algorithm becomes reducible. The methodology developed in Roberts and Stramer [2001a. On inference for partial observed nonlinear diffusion models using the metropolis-hastings algorithm. Biometrika 88(3); 603–621] circumvents the problem for scalar diffusions. We extend this framework to the class of models of this paper by introducing an appropriate reparametrisation of the likelihood that can be used to construct an irreducible data augmentation scheme. Practical implementation issues are considered and the methodology is applied to simulated data from the Heston model.     

Transient probability functions of finite birth–death processes with catastrophes

A representation of the transient probability functions of finite birth–death processes (with or without catastrophes) as a linear combination of exponential functions is derived using a recursive, Cayley–Hamilton approach. This method of solution allows practitioners to solve for these transient probability functions by reducing the problem to three calculations: determining eigenvalues of the Q$Q$-matrix, raising the Q$Q$-matrix to an integer power and solving a system of linear equations. The approach avoids Laplace transforms and permits solution of a particular transition probability function from state i$i$ to j$j$ without determining all such functions.     

A comparison between several adjustment models to simulated teletraffic data

In this paper we assess the suitability of the fractional Brownian motion and the Markovian modulated Poisson process to represent modern teletraffic data when we use them in dimensioning and quality of service studies. The procedure consists in adjusting a model of each family to a stream of data simulated with a physical model that emulates the real flow of data. Then, both models are profusely studied in order to establish which properties of the original data are captured.     

Birth-multiple catastrophe processes

Birth-multiple catastrophe processes are analyzed where the birth transition rates are assumed to be constant while catastrophes are distinguished by having possibly different destinations and possibly different transition rates. The transient probability functions of such birth-multiple catastrophe systems are determined. The solution method uses dual processes, randomization, and sample path counting. Solutions are explicit in terms of being a finite linear combination of products of exponential functions of time, t, and nonnegative integer powers of t. The coefficients within this expansion follow a pattern of rational functions of the transition rates.     

Stochastic volatility modelling in continuous time with general marginal distributions: Inference,prediction and model selection

We compare results for stochastic volatility models where the underlying volatility process having generalized inverse Gaussian (GIG) and tempered stable marginal laws. We use a continuous time stochastic volatility model where the volatility follows an Ornstein–Uhlenbeck stochastic differential equation driven by a Lévy process. A model for long-range dependence is also considered, its merit and practical relevance discussed. We find that the full GIG and a special case, the inverse gamma, marginal distributions accurately fit real data. Inference is carried out in a Bayesian framework, with computation using Markov chain Monte Carlo (MCMC). We develop an MCMC algorithm that can be used for a general marginal model.     

A generalized drop-the-loser rule for multi-treatment clinical trials

Urn models are popular for response adaptive designs in clinical studies. Among different urn models, Ivanova's drop-the-loser rule is capable of producing superior adaptive treatment allocation schemes. Ivanova [2003. A play-the-winner-type urn model with reduced variability. Metrika 58, 1–13] obtained the asymptotic normality only for two treatments. Recently, Zhang et al. [2007. Generalized drop-the-loser urn for clinical trials with delayed responses. Statist. Sinica, in press] extended the drop-the-loser rule to tackle more general circumstances. However, their discussion is also limited to only two treatments. In this paper, the drop-the-loser rule is generalized to multi-treatment clinical trials, and delayed responses are allowed. Moreover, the rule can be used to target any desired pre-specified allocation proportion. Asymptotic properties, including strong consistency and asymptotic normality, are also established for general multi-treatment cases.     

On quadratic prediction problems in finite populations

For a finite population and its linear model, Liu and Rong proposed a notion of optimal invariant quadratic unbiased prediction (OIQUP) and offered two methods for studying this notion, in which the first is incomplete. In this note, we mainly aim at fulfilling the first approach used by Liu and Rong by considering a transformed matrix equation set through permutation matrix techniques. Solvability of the matrix equation set, optimality of the resulting predictor, and equivalence of the representations of OIQUP, derived in this note and by Liu and Rong, are investigated in detail. In addition, an application to predicting population variance is conducted based on a simulated population.     

Ordering conditional lifetimes of coherent systems

Consider a system of n components that has the property that there exists a number r  (r<n)$(r<n)$, such that if it is known that at most r components have failed, the system is still functioning with probability 1. Suppose that such a system is equipped with a warning light that comes up at the time of the failure of the rth component. The system is still working then, and we are interested in its residual life. In this paper we obtain some results which stochastically compare the residual lives of such systems with the same type, or with different types, of components. Some applications are given. In particular, we derive upper and lower bounds on the expected residual lives of such systems given that the warning light has not come up yet, and given that the component hazard rate functions are bounded from below or from above by a known constant.     

Likelihood ratio ordering of order statistics,mixtures and systems

Let X=(_X1,_X2,…,_Xn)$\mathit{X}=(X_1,X_2,\dots ,X_n)$ be an exchangeable random vector, and denote _X1:i=min{_X1,_X2,…,_Xi}$X_{1:i}=\min \{X_1,X_2,\dots ,X_i\}$ and _Xi:i=max{_X1,_X2,…,_Xi}$X_{i:i}=\max \{X_1,X_2,\dots ,X_i\}$, 1?i?n$1?i?n$. These order statistics represent the lifetimes of the series and the parallel systems, respectively, with component lifetimes _Xi$X_i$. In this paper we obtain conditions under which _X1:i$X_{1:i}$ (or _Xi:i$X_{i:i}$) decreases (increases) in i in the likelihood ratio (lr) order. An even more general result involving general (that is, not necessary exchangeable) random vectors is also derived for general series (or parallel) systems. We show that the series (parallel) systems are not necessarily lr-ordered even if the components are independent.     

Occupancy numbers for dynamic heterogeneous populations: Estimate of particles lifetimes

A dynamic model of a heterogeneous population is studied. Particles belonging to a population are divided, at every time t, into a finite number of classes according to their types and the partition changes over time. The role of the occupancy numbers, namely the cardinality of each class, is highlighted. The relationship between the stochastic process of occupancy numbers and the process of particle types is analyzed. The main goal of this paper is the estimation of the lifetime of each particle at a given time t, when the observed data are the history of the process of the number of dead particles up to t. Furthermore, a discrete time approximation of the filter is given.