VARIATIONAL BAYESIAN ANALYSIS FOR HIDDEN MARKOV MODELS

The variational approach to Bayesian inference enables simultaneous estimation of model parameters and model complexity. An interesting feature of this approach is that it also leads to an automatic choice of model complexity. Empirical results from the analysis of hidden Markov models with Gaussian observation densities illustrate this. If the variational algorithm is initialized with a large number of hidden states, redundant states are eliminated as the method converges to a solution, thereby leading to a selection of the number of hidden states. In addition, through the use of a variational approximation, the deviance information criterion for Bayesian model selection can be extended to the hidden Markov model framework. Calculation of the deviance information criterion provides a further tool for model selection, which can be used in conjunction with the variational approach.     

Palm Distributions for Log Gaussian Cox Processes

This paper establishes a remarkable result regarding Palm distributions for a log Gaussian Cox process: the reduced Palm distribution for a log Gaussian Cox process is itself a log Gaussian Cox process that only differs from the original log Gaussian Cox process in the intensity function. This new result is used to study functional summaries for log Gaussian Cox processes.     

Estimation of P{X < Y} for geometric—exponential model based on complete and censored samples

This article deals with the estimation of R = P{X < Y}, where X and Y are independent random variables from geometric and exponential distribution, respectively. For complete samples, the MLE of R, its asymptotic distribution, and confidence interval based on it are obtained. The procedure for deriving bootstrap-p confidence interval is presented. The UMVUE of R and UMVUE of its variance are derived. The Bayes estimator of R is investigated and its Lindley's approximation is obtained. A simulation study is performed in order to compare these estimators. Finally, all point estimators for right censored sample from the exponential distribution, are obtained.     

Saddlepoint p-values for group of linear rank tests

Many nonparametric tests in one sample problem, matched pairs, and competingrisks under censoring have the same underlying permutation distribution. This article proposes a saddlepoint approximation to the exact p-values of these tests instead of the asymptotic approximations. The performance of the saddlepoint approximation is assessed by using simulation studies that show the superiority of the saddlepoint methods over the asymptotic approximations in several settings. The use of the saddlepoint to approximate the p-values of class of two sample tests under complete randomized design is also discussed.     

Recovery of functions from transformed moments: A unified approach

Two approximations recovering the functions from their transformed moments are proposed. The upper bounds for the uniform rate of convergence are derived. In addition, the comparisons of the estimates of the cumulative distribution function and its density function with the empirical distribution and the kernel density estimates are conducted via a simulation study. The plots of recovered functions are presented for several examples as well.     

Retrospective change detection in categorical time series

The aim of this paper is to propose methods of detecting change in the coefficients of a multinomial logistic regression model for categorical time series offline. The alternatives to the null hypothesis of stationarity can be either the hypothesis that it is not true, or that there is a temporary change in the sequence. We use the efficient score vector of the partial likelihood function. This has several advantages. First, the alternative value of the parameter does not have to be estimated; hence, we have a procedure that has a simple structure with only one parameter estimation using all available observations. This is in contrast with the generalized likelihood ratio-based change point tests. The efficient score vector is used in various ways. As a vector, its components correspond to the different components of the multinomial logistic regression model’s parameter vector. Using its quadratic form a test can be defined, where the presence of a change in any or all parameters is tested for. If there are too many parameters one can test for any subset while treating the rest as nuisance parameters. Our motivating example is a DNA sequence of four categories, and our test result shows that in the published data the distribution of the four categories is not stationary.     

一种基于加权F-范数的半正定矩阵的逼近方法

正定性是许多金融预测模型的重要假设前提，然而从实际样本中得到的相关系数矩阵并不能保证其正定性。为此在介绍如何根据样本设定相关系数矩阵以及范数逼近原理的基础上，如何根据该原理找到与之最接近的相关系数矩阵，即最接近的单位对角半正定对称矩阵。通过实证，验证了其方法的有效性。     

非线性回归分析的一种有理逼近

利用有理插值逼近回归函数,把最小二乘法和插值技术结合起来,充分利用已知信息,达到最佳逼近,通过实例,获得了很好的效果。     

A probabilistic approach to Wallis’ formula

We present an alternative proof of Wallis’ formula from the probabilistic point of view. Based on the classical central limit theorem, some discrete distributions with additive property, such as binomial, negative binomial, Poisson and multinomial distributions, are considered to derive π/2.     

Evaluation of the parameters of su by rational fractions

A rational fraction approximation is given for a function of one of the parameters defining Johnson's S_UError assessment for a segment of the domain of validity shows remarkable accuracy.