Objective Bayesian analysis for the accelerated degradation model based on the inverse Gaussian process

The inverse Gaussian process is an attractive stochastic process to model monotone degradation paths in degradation analysis. In this paper, we propose an objective Bayesian method to analyze the accelerated degradation model based on the inverse Gaussian process. Noninformative priors including the Jeffreys prior and reference priors are derived, and the propriety of the posteriors under each prior is validated. A simulation study is carried out to investigate the performance of the approach compared with the maximum likelihood method and the Bootstrap method. Numerical results show that the proposed method has better performance in terms of the mean squared error and the frequentist coverage probability. The reference prior ${\pi }_{R_2}$ is recommended to use in practice. Finally, the Bayesian approach is applied to a real data.     

Bayesian normal analysis with an inverse Gaussian prior

Summary This paper considers the Bayesian analysis of normal distribution when its variance has an inverse Gaussian prior density. The result is also generelized in a theorem that is subsequently presented.     

The transformed inverse Gaussian process as an age- and state-dependent degradation model

In this paper, a transformed inverse Gaussian (TIG) process is introduced as a new family of monotonic degradation models. Different from most state-of-the-art degradation models, which can only characterize age-dependent performance degradation, the TIG process model is mainly introduced for degradation modelling of industrial products with age- and state-dependent performance degradation. With this new model, promising properties include (1) the modelling capability for characterizing products observed at discrete time points with age- and state-dependent degradation, (2) the mathematical tractability for calculating the reliability function and remaining useful life distribution with high efficiency, and (3) the modelling flexibility of incorporating explanatory variables and random effects for investigating a product population with unit-to-unit heterogeneity. To facilitate the degradation modelling and analysis, methods for parameter estimation and model selection are developed under a coherent Bayesian framework. Simulation studies and real cases are presented to demonstrate the proposed degradation model and the Bayesian methods.     

Bayesian nonstationary Gaussian process models via treed process convolutions

Advances in Data Analysis and Classification - The Gaussian process is a common model in a wide variety of applications, such as environmental modeling, computer experiments, and geology. Two major...     

Bayesian reference analysis for Gaussian Markov random fields

Gaussian Markov random fields (GMRF) are important families of distributions for the modeling of spatial data and have been extensively used in different areas of spatial statistics such as disease mapping, image analysis and remote sensing. GMRFs have been used for the modeling of spatial data, both as models for the sampling distribution of the observed data and as models for the prior of latent processes/random effects; we consider mainly the former use of GMRFs. We study a large class of GMRF models that includes several models previously proposed in the literature. An objective Bayesian analysis is presented for the parameters of the above class of GMRFs, where explicit expressions for the Jeffreys (two versions) and reference priors are derived, and for each of these priors results on posterior propriety of the model parameters are established. We describe a simple MCMC algorithm for sampling from the posterior distribution of the model parameters, and study frequentist properties of the Bayesian inferences resulting from the use of these automatic priors. Finally, we illustrate the use of the proposed GMRF model and reference prior for studying the spatial variability of lip cancer cases in the districts of Scotland over the period 1975-1980.     

Degradation modeling with subpopulation heterogeneities based on the inverse Gaussian process

This study proposes a random effects model based on inverse Gaussian process, where the mixture normal distribution is used to account for both unit-specific and subpopulation-specific heterogeneities. The proposed model can capture heterogeneities due to subpopulations in the same population or the units from different batches. A new Expectation-Maximization (EM) algorithm is developed for point estimation and the bias-corrected bootstrap is used for interval estimation. We show that the EM algorithm updates the parameters based on the gradient of the loglikelihood function via a projection matrix. In addition, the convergence rate depends on the condition number that can be obtained by the projection matrix and the Hessian matrix of the loglikelihood function. A simulation study is conducted to assess the proposed model and the inference methods, and two real degradation datasets are analyzed for illustration.     

A semi-Markov process with an inverse Gaussian distribution as sojourn time

Let X_n denote the state of a device after n repairs. We assume that the time between two repairs is the time τ taken by a Wiener process {W(t), t ? 0}, starting from w₀ and with drift μ < 0, to reach c ∈ [0, w₀). After the nth repair, the process takes on either the value X_n?1 + 1 or X_n?1 + 2. The probability that X_n = X_n?1 + j, for j = 1, 2, depends on whether τ ? t₀ (a fixed constant) or τ > t₀. The device is considered to be worn out when X_n ? k, where k ∈ {1, 2, …}. This model is based on the ones proposed by Rishel (1991) [1] and Tseng and Peng (2007) [2]. We obtain an explicit expression for the mean lifetime of the device. Numerical methods are used to illustrate the analytical findings.     

A perturbed gamma degradation process with degradation dependent non‐Gaussian measurement errors

This paper proposes and illustrates a new perturbed gamma degradation process where the measurement error is modeled as a non‐Gaussian random variable that depends stochastically on the actual degradation level. The expression of the likelihood function for a generic set of noisy degradation measurements is derived, and the expression of the remaining useful life distribution of a degrading unit that fails when its degradation level exceeds a given threshold limit is formulated. A particle filter method is suggested, which allows one to compute the likelihood function and to estimate the remaining useful life distribution in a quick yet efficient manner. In addition, a closed‐form approximation of the perturbed gamma process is proposed to use in the special, yet meaningful, case where the standard deviation of the measurement error depends linearly on the actual degradation level. Finally, an applicative example is discussed, where the parameters of the perturbed gamma process, the remaining useful life distribution, and the mean remaining useful life of the degrading units are estimated from a set of noisy real degradation data.     

Bayesian estimation and prediction for the transformed Wiener degradation process

This paper proposes some Bayesian inferential procedures for the transformed Wiener (TW) process, a new degradation process that has been recently suggested in the literature to describe degradation phenomena where degradation increments are not necessarily positive and depend stochastically on the current degradation level. These procedures have been expressly conceived to allow one incorporating into the inferential process the type of prior information, on meaningful physical characteristics of the observed degradation process, that is generally available in practical settings. Several different prior distributions are proposed, each of them reflecting a specific degree of knowledge on the observed phenomenon. Simple strategies for eliciting the prior hyper‐parameters from the available prior information are provided. Estimates of the TW process parameters and some functions thereof are retrieved by adopting a Monte Carlo Markov Chain technique. Procedures that allow predicting the degradation increment, the useful life of a new unit, and the remaining useful life of a used unit are also provided. Finally, an application is developed on the basis of a set of real degradation measurements of some infrared light‐emitting diodes, widely used in communication systems. The obtained results demonstrate the feasibility of the proposed Bayesian approach and the flexibility of the TW process.     

Fractional normal inverse Gaussian diffusion

A fractional normal inverse Gaussian (FNIG) process is a fractional Brownian motion subordinated to an inverse Gaussian process. This paper shows how the FNIG process emerges naturally as the limit of a random walk with correlated jumps separated by i.i.d. waiting times. Similarly, we show that the NIG process, a Brownian motion subordinated to an inverse Gaussian process, is the limit of a random walk with uncorrelated jumps separated by i.i.d. waiting times. The FNIG process is also derived as the limit of a fractional ARIMA processes. Finally, the NIG densities are shown to solve the relativistic diffusion equation from statistical physics.     

Score tests for inverse Gaussian mixtures

The mixed inverse Gaussian given by Whitmore (biScand. J. Statist., 13 , 1986, 211–220) provides a convenient way for testing the goodness‐of‐fit of a pure inverse Gaussian distribution. The test is a one‐sided score test with the null hypothesis being the pure inverse Gaussian (i.e. the mixing parameter is zero) and the alternative a mixture. We devise a simple score test and study its finite sample properties. Monte Carlo results show that it compares favourably with the smooth test of Ducharme ( Test , 10 , 2001, 271‐290). In practical applications, when the pure inverse Gaussian distribution is rejected, one is interested in making inference about the general values of the mixing parameter. However, as it is well known that the inverse Gaussian mixture is a defective distribution; hence, the standard likelihood inference cannot be applied. We propose several alternatives and provide score tests for the mixing parameter. Finite sample properties of these tests are examined by Monte Carlo simulation. Copyright © 2010 John Wiley & Sons, Ltd.     

The exponentiated generalized inverse Gaussian distribution

The modeling and analysis of lifetime data is an important aspect of statistical work in a wide variety of scientific and technological fields. Good (1953) introduced a probability distribution which is commonly used in the analysis of lifetime data. For the first time, based on this distribution, we propose the so-called exponentiated generalized inverse Gaussian distribution, which extends the exponentiated standard gamma distribution (Nadarajah and Kotz, 2006). Various structural properties of the new distribution are derived, including expansions for its moments, moment generating function, moments of the order statistics, and so forth. We discuss maximum likelihood estimation of the model parameters. The usefulness of the new model is illustrated by means of a real data set.     

Adaptive Gaussian process emulators for efficient reliability analysis

This paper presents an approximation method for performing efficient reliability analysis with complex computer models. The computational cost of industrial-scale models can cause problems when performing sampling-based reliability analysis. This is due to the fact that the failure modes of the system typically occupy a small region of the performance space and thus require relatively large sample sizes to accurately estimate their characteristics. The sequential sampling method proposed in this article, combines Gaussian process-based optimisation and subset simulation. Gaussian process emulators construct a statistical approximation to the output of the original code, which is both affordable to use and has its own measure of predictive uncertainty. Subset simulation is used as an integral part of the algorithm to efficiently populate those regions of the surrogate which are likely to lead to the performance function exceeding a predefined critical threshold. The emulator itself is used to inform decisions about efficiently using the original code to augment its predictions. The iterative nature of the method ensures that an arbitrarily accurate approximation of the failure region is developed at a reasonable computational cost. The presented method is applied to an industrial model of a biodiesel filter.     

Algebraic geometry of Gaussian Bayesian networks

Conditional independence models in the Gaussian case are algebraic varieties in the cone of positive definite covariance matrices. We study these varieties in the case of Bayesian networks, with a view towards generalizing the recursive factorization theorem to situations with hidden variables. In the case when the underlying graph is a tree, we show that the vanishing ideal of the model is generated by the conditional independence statements implied by graph. We also show that the ideal of any Bayesian network is homogeneous with respect to a multigrading induced by a collection of upstream random variables. This has a number of important consequences for hidden variable models. Finally, we relate the ideals of Bayesian networks to a number of classical constructions in algebraic geometry including toric degenerations of the Grassmannian, matrix Schubert varieties, and secant varieties.     

Bayesian Networks with Degenerate Gaussian Distributions

Bayesian networks compute marginal distributions through the message passing algorithm—a series of local calculations involving neighboring cliques of variables in a clique tree. In this work, these message passing computations are extended to the case of degenerate Gaussian potentials which are multivariate Gaussian densities that can have two different kinds of degeneracies corresponding to projections with zero variance and projections with infinite variance. The basic operations of the message passing algorithm, such as representing conditional distributions, extending potentials, and conditioning on observations, create or operate on potentials with various kinds of degeneracies thereby demonstrating the need for such methodology. Computer implementation of this scheme follows easily from the details within and some computational aspects are discussed. We also demonstrate an application of our methodology to automatic musical accompaniment.     

Extreme inaccuracies in Gaussian Bayesian networks

To evaluate the impact of model inaccuracies over the network’s output, after the evidence propagation, in a Gaussian Bayesian network, a sensitivity measure is introduced. This sensitivity measure is the Kullback-Leibler divergence and yields different expressions depending on the type of parameter to be perturbed, i.e. on the inaccurate parameter.In this work, the behavior of this sensitivity measure is studied when model inaccuracies are extreme, i.e. when extreme perturbations of the parameters can exist. Moreover, the sensitivity measure is evaluated for extreme situations of dependence between the main variables of the network and its behavior with extreme inaccuracies. This analysis is performed to find the effect of extreme uncertainty about the initial parameters of the model in a Gaussian Bayesian network and about extreme values of evidence. These ideas and procedures are illustrated with an example.     

Asymptotics and bootstrap for inverse Gaussian regression

This paper studies regression, where the reciprocal of the mean of a dependent variable is considered to be a linear function of the regressor variables, and the observations on the dependent variable are assumed to have an inverse Gaussian distribution. The large sample theory for the pseudo maximum likelihood estimators is available in the literature, only when the number of replications increase at a fixed rate. This is inadequate for many practical applications. This paper establishes consistency and derives the asymptotic distribution for the pseudo maximum likelihood estimators under very general conditions on the design points. This includes the case where the number of replications do not grow large, as well as the one where there are no replications. The bootstrap procedure for inference on the regression parameters is also investigated.Research supported in part by NSF Grant DMS-9208066.Research supported in part by NSERC of Canada.     

Bayes estimation in a mixture inverse Gaussian model

In this paper a mixture model involving the inverse Gaussian distribution and its length biased version is studied from a Bayesian view-point. Using proper priors, the Bayes estimates of the parameters of the model are derived and the results are applied on the aircraft data of Proschan (1963,Technometrics,5, 375–383). The posterior distributions of the parameters are expressed in terms of the confluent-hypergeometric function and the modified Bessel function of the third kind. The integral involved in the expression of the estimate of the mean is evaluated by numerical techniques.     

Simultaneous confidence intervals for several inverse Gaussian populations

In this research, we propose simultaneous confidence intervals for all pairwise comparisons of means from inverse Gaussian distribution. Our method is based on fiducial generalized pivotal quantities for vector parameters. We prove that the constructed confidence intervals have asymptotically correct coverage probabilities. Simulation results show that the simulated Type-I errors are close to the nominal level even for small samples. The proposed approach is illustrated by an example.