Conditional failure occurrence rates for semi-Markov chains*

ABSTRACT

In the present paper, we aim at providing plug-in-type empirical estimators that enable us to quantify the contribution of each operational or/and non-functioning state to the failures of a system described by a semi-Markov model. In the discrete-time and finite state space semi-Markov framework, we study different conditional versions of an important reliability measure for random repairable systems, the failure occurrence rate, which is based on counting processes. The identification of potential failure contributors through the conditional counterparts of the failure occurrence rate is of paramount importance since it could lead to corrective actions that minimize the occurrence of the more important failure modes and therefore improve the reliability of the system. The aforementioned estimators are characterized by appealing asymptotic properties such as strong consistency and asymptotic normality. We further obtain detailed analytical expressions for the covariance matrices of the random vectors describing the conditional failure occurrence rates. As particular cases we present the failure occurrence rates for hidden (semi-) Markov models. We illustrate our results by means of a simulated study. Different applications are presented based on wind, earthquake and vibration data.     

Asymptotic theory for the Cox semi-Markov illness-death model

Irreversible illness-death models are used to model disease processes and in cancer studies to model disease recovery. In most applications, a Markov model is assumed for the multistate model. When there are covariates, a Cox (1972, J Roy Stat Soc Ser B 34:187–220) model is used to model the effect of covariates on each transition intensity. Andersen et al. (2000, Stat Med 19:587–599) proposed a Cox semi-Markov model for this problem. In this paper, we study the large sample theory for that model and provide the asymptotic variances of various probabilities of interest. A Monte Carlo study is conducted to investigate the robustness and efficiency of Markov/Semi-Markov estimators. A real data example from the PROVA (1991, Hepatology 14:1016–1024) trial is used to illustrate the theory.     

Likelihood-based estimation for Gaussian MRFs

Markov random fields (MRFs) express spatial dependence through conditional distributions, although their stochastic behavior is defined by their joint distribution. These joint distributions are typically difficult to obtain in closed form, the problem being a normalizing constant that is a function of unknown parameters. The Gaussian MRF (or conditional autoregressive model) is one case where the normalizing constant is available in closed form; however, when sample sizes are moderate to large (thousands to tens of thousands), and beyond, its computation can be problematic. Because the conditional autoregressive (CAR) model is often used for spatial-data modeling, we develop likelihood-inference methodology for this model in situations where the sample size is too large for its normalizing constant to be computed directly. In particular, we use simulation methodology to obtain maximum likelihood estimators of mean, variance, and spatial-depencence parameters (including their asymptotic variances and covariances) of CAR models.     

Markov Systems in a General State Space

In this article, we introduce and study Markov systems on general spaces (MSGS) as a first step of an entire theory on the subject. Also, all the concepts and basic results needed for this scope are given and analyzed. This could be thought of as an extension of the theory of a non homogeneous Markov system (NHMS) and that of a non homogeneous semi-Markov system on countable spaces, which has realized an interesting growth in the last thirty years. In addition, we study the asymptotic behaviour or ergodicity of Markov systems on general state spaces. The problem of asymptotic behaviour of Markov chains has been central for finite or countable spaces since the foundation of the subject. It has also been basic in the theory of NHMS and NHSMS. Two basic theorems are provided in answering the important problem of the asymptotic distribution of the population of the memberships of a Markov system that lives in the general space (X, ?(X)). Finally, we study the total variability from the invariant measure of the Markov system given that there exists an asymptotic behaviour. We prove a theorem which states that the total variation is finite. This problem is known also as the coupling problem.     

Asymptotic distributions of functions of a sample covariance matrix under the elliptical distribution

This paper is concerned with asymptotic distributions of functions of a sample covariance matrix under the elliptical model. Simple but useful formulae for calculating asymptotic variances and covariances of the functions are derived. Also, an asymptotic expansion formula for the expectation of a function of a sample covariance matrix is derived; it is given up to the second-order term with respect to the inverse of the sample size. Two examples are given: one of calculating the asymptotic variances and covariances of the stepdown multiple correlation coefficients, and the other of obtaining the asymptotic expansion formula for the moments of sample generalized variance.     

Exotic Properties of Non Homogeneous Markov and Semi-Markov Systems

In this article we study what we chose to call exotic properties of NHMS and NHSMS. The interplay between stochastic theory of NHMS and NHSMS and other branches of probability, stochastic processes and mathematics, we believe is a fascinating one apart from being important. In many cases the information needed for the evolution of a NHMS is a larger set than the history of the multidimensional process NHMS. In our world where an overflow of information exists almost in all problems, it is almost surely that this will be available. Here, we extend the definition of the NHMS in order to accomodate this case. In this respect we arrive at the defnition of the 𝒢-non homogeneous Markov system. We study the problem of change of measure in a 𝒢-non homogeneousMarkov system. It is proved that under certain conditions the NHMS retains the Markov property, while as expected the basic sequences of transition probabilities change and it is established how they do so. We also find the expected population structure of the NHMS under the new measure in close analytic form. We also define the 𝒢-non homogeneous semi-Markov system and we study the problem of change of measure in a 𝒢-non homogeneous semi-Markov system. It is proved that under certain conditions the NHSMS retains the semi-Markov property while as expected the basic sequences of transition probabilities change and it is established how they do so. We prove that if the input process of memberships is a non homogeneous Poisson process, then asymtotically and under certain easily met in practice conditions, the compensated population structure of the 𝒢-NHMS is a martingale. Finally we prove that the space of all random population structures, under easily met in practice conditions, is a Hilbert space.     

Problems with Maximum Likelihood Estimation and the 3 Parameter Gamma Distribution

The three parameters involved are scale a , shape 𝜌 , and location s . Maximum likelihood estimators are (\hata, \hat\rho, \hats) . Using recent work on the second order variances, skewness, and kurtosis we establish the facts, that if the location parameter s is to be estimated, then the asymptotic variances only exist if 𝜌 >2, asymptotic skewness only exists if 𝜌 >3, and 2nd order variances and third order fourth central moments only exist if 𝜌 >4. The result of these limitations is that in general very large sample sizes may be needed to avoid inference problems. We also include new continued fractions for the asymptotic covariances of the maximum likelihood estimators considered.     

Bivariate Semi-Markov Process for Counterparty Credit Risk

We consider the problem of constructing an appropriate multivariate model to study counterparty credit risk in the credit rating migration problem. For this financial problem different multivariate Markov chain models were proposed. However, the Markovian assumption may be inappropriate for the study of the dynamics of credit ratings, which typically show non Markovian-like behavior. In this article, we develop a semi-Markov approach to study the counterparty credit risk by defining a new multivariate semi-Markov chain model. Methods are given for computing the transition probabilities, reliability functions and the price of a risky Credit Default Swap.     

Nonparametric test for the homogeneity of the overall variability

In this paper, we propose a nonparametric test for homogeneity of overall variabilities for two multi-dimensional populations. Comparisons between the proposed nonparametric procedure and the asymptotic parametric procedure and a permutation test based on standardized generalized variances are made when the underlying populations are multivariate normal. We also study the performance of these test procedures when the underlying populations are non-normal. We observe that the nonparametric procedure and the permutation test based on standardized generalized variances are not as powerful as the asymptotic parametric test under normality. However, they are reliable and powerful tests for comparing overall variability under other multivariate distributions such as the multivariate Cauchy, the multivariate Pareto and the multivariate exponential distributions, even with small sample sizes. A Monte Carlo simulation study is used to evaluate the performance of the proposed procedures. An example from an educational study is used to illustrate the proposed nonparametric test.     

Approximations of Variances and Covariances for Order Statistics from the Standard Extreme Value Distribution

We consider simple approximations of variances and covariances for order statistics from the standard extreme value distribution. Exact values and simulation results of the variances and covariances for certain sample sizes are used to determine the validity of the suggested approximations.     

First hitting probabilities for semi-Markov chains and estimation

We first consider a stochastic system described by an absorbing semi-Markov chain (SMC) with finite state space, and we introduce the absorption probability to a class of recurrent states. Afterwards, we study the first hitting probability to a subset of states for an irreducible SMC. In the latter case, a non-parametric estimator for the first hitting probability is proposed and the asymptotic properties of strong consistency and asymptotic normality are proven. Finally, a numerical application on a five-state system is presented to illustrate the performance of this estimator.     

Higher Order Asymptotic Cumulants of Studentized Estimators in Covariance Structures

In covariance structure analysis, the Studentized pivotal statistic of a parameter estimator is often used since the statistic is asymptotically normally distributed with mean zero and unit variance. For more accurate asymptotic distribution, the first and third asymptotic cumulants can be used to have the single-term Edgeworth, Cornish-Fisher, and Hall type asymptotic expansions. In this paper, the higher order asymptotic variance and the fourth asymptotic cumulant of the statistic are obtained under nonnormality when the partial derivatives of a parameter estimator with respect to sample variances and covariances up to the third order and the moments of the associated observed variables up to the eighth order are available. The result can be used to have the two-term Edgeworth expansion. Simulations are performed to see the accuracy of the asymptotic results in finite samples.     

On Reproducing Linear Estimators within the Gauss–Markov Model with Stochastic Constraints

In a Gauss–Markov Model (GMM) with fixed constraints, all the relevant estimators perfectly satisfy these constraints. As soon as they become stochastic, most estimators are allowed to satisfy them only approximately, thereby leaving room for nonvanishing residuals to describe the deviation from the prior information.

Sometimes, however, linear estimators may be preferred that are able to perfectly reproduce the prior information in form of stochastic constraints, including their variances and covariances. As typical example may be considered the case where a geodetic network ought to be densified without changing the higher-order point coordinates that are usually introduced together with their variances and (some) covariances. Traditional estimators are based on the “Helmert” or “S-transformation,” respectively an adaptation of the fixed-constraints Least-Squares estimator.

Here we show that neither approach generates the optimal reproducing estimator, which will be presented in detail and compared with the other reproducing estimators in terms of their MSE-risks.     

Local composite quantile regression estimation of time-varying parameter vector for multidimensional time-inhomogeneous diffusion models

This paper is dedicated to the study of the composite quantile regression (CQR) estimations of time-varying parameter vectors for multidimensional diffusion models. Based on the local linear fitting for parameter vectors, we propose the local linear CQR estimations of the drift parameter vectors, and verify their asymptotic biases, asymptotic variances and asymptotic normality. Moreover, we discuss the asymptotic relative efficiency (ARE) of the local linear CQR estimations with respect to the local linear least-squares estimations. We obtain that the local estimations that we proposed are much more efficient than the local linear least-squares estimations. Simulation studies are constructed to show the performance of the estimations proposed.     

基于体制转换模型的动态相关系数研究

对相关系数的马尔可夫体制转换模型（RSDC）进行拓展,将多维随机变量的协方差阵分成方差和相关系数,方差利用GARCH模型来描述,相关系数阵则划分为不同的状态。RSDC模型被应用于中国股票市场研究板块指数的动态相关性,给出行业板块的beta系数。实证结果表明两状态恰好体现高、低不同的相关性,且状态持续的概率较大,基于RSDC模型的beta系数与常相关系数的多元GARCH模型存在一定的差异。     

Large Deviations for Empirical Estimators of the Stationary Distribution of a Semi-Markov Process with Finite State Space

We prove the large deviation principle for empirical estimators of stationary distributions of semi-Markov processes with finite state space, irreducible embedded Markov chain, and finite mean sojourn time in each state. We consider on/off Gamma sojourn processes as an illustrative example, and, in particular, continuous time Markov chains with two states. In the second case, we compare the rate function in this article with the known rate function concerning another family of empirical estimators of the stationary distribution.     

Reducing Periodicity of Fuzzy Markov Chains Based on Simulation Using Halton Sequence

We first introduce fuzzy finite Markov chains and present some of their fundamental properties based on possibility theory. We also bring in a way to convert fuzzy Markov chains to classic Markov chains. In addition, we simulate fuzzy Markov chain using different sizes. It is observed that the most of fuzzy Markov chains not only do have an ergodic behavior, but also they are periodic. Finally, using Halton quasi-random sequence we generate some fuzzy Markov chains, which are compared with the ones generated by the RAND function of MATLAB. Therefore, we improve the periodicity behavior of fuzzy Markov chains.     

A Note on the Canonical Correlations From Contingency Tables

Explicit formulae are obtained for the asymptotic variances and covariances of canonical correlations which correspond to non-zero theoretical correlations in (p+ 1) x (q+1) contingency tables, with p≤q. The moments of the roots of a central Wishart matrix distributed as Wp(q; I ) are also given in general, with means, variances and covariances tabulated for p= 2, 3, 4: these may apply to canonical correlations corresponding to zeros.     

Recursive computation of the single and product moments of order statistics from the complementary exponential–geometric distribution

The complementary exponential–geometric distribution has been proposed recently as a simple and useful reliability model for analysing lifetime data. For this distribution, some recurrence relations are established for the single and product moments of order statistics. These recurrence relations enable the computation of the means, variances and covariances of all order statistics for all sample sizes in a simple and efficient recursive manner. By using these relations, we have tabulated the means, variances and covariances of order statistics from samples of sizes up to 10 for various values of the shape parameter θ. These values are in turn used to determine the best linear unbiased estimator of the scale parameter β based on complete and Type-II right-censored samples.