Computing marginal likelihoods from a single MCMC output

In this article, we propose new Monte Carlo methods for computing a single marginal likelihood or several marginal likelihoods for the purpose of Bayesian model comparisons. The methods are motivated by Bayesian variable selection, in which the marginal likelihoods for all subset variable models are required to compute. The proposed estimates use only a single Markov chain Monte Carlo (MCMC) output from the joint posterior distribution and it does not require the specific structure or the form of the MCMC sampling algorithm that is used to generate the MCMC sample to be known. The theoretical properties of the proposed method are examined in detail. The applicability and usefulness of the proposed method are demonstrated via ordinal data probit regression models. A real dataset involving ordinal outcomes is used to further illustrate the proposed methodology.     

A Bayesian approach to bandwidth selection for multivariate kernel regression with an application to state-price density estimation

This paper presents a Bayesian approach to bandwidth selection for multivariate kernel regression. A Monte Carlo study shows that under the average squared error criterion, the Bayesian bandwidth selector is comparable to the cross-validation method and clearly outperforms the bootstrapping and rule-of-thumb bandwidth selectors. The Bayesian bandwidth selector is applied to a multivariate kernel regression model that is often used to estimate the state-price density of Arrow–Debreu securities with the S&P 500 index options data and the DAX index options data. The proposed Bayesian bandwidth selector represents a data-driven solution to the problem of choosing bandwidths for the multivariate kernel regression involved in the nonparametric estimation of the state-price density pioneered by Aït-Sahalia and Lo [Aït-Sahalia, Y., Lo, A.W., 1998. Nonparametric estimation of state-price densities implicit in financial asset prices. The Journal of Finance, 53, 499, 547.]     

Bayesian median autoregression for robust time series forecasting

We develop a Bayesian median autoregressive (BayesMAR) model for time series forecasting. The proposed method utilizes time-varying quantile regression at the median, favorably inheriting the robustness of median regression in contrast to the widely used mean-based methods. Motivated by a working Laplace likelihood approach in Bayesian quantile regression, BayesMAR adopts a parametric model bearing the same structure as autoregressive models by altering the Gaussian error to Laplace, leading to a simple, robust, and interpretable modeling strategy for time series forecasting. We estimate model parameters by Markov chain Monte Carlo. Bayesian model averaging is used to account for model uncertainty, including the uncertainty in the autoregressive order, in addition to a Bayesian model selection approach. The proposed methods are illustrated using simulations and real data applications. An application to U.S. macroeconomic data forecasting shows that BayesMAR leads to favorable and often superior predictive performance compared to the selected mean-based alternatives under various loss functions that encompass both point and probabilistic forecasts. The proposed methods are generic and can be used to complement a rich class of methods that build on autoregressive models.     

Time-varying sparsity in dynamic regression models

A novel Bayesian method for inference in dynamic regression models is proposed where both the values of the regression coefficients and the importance of the variables are allowed to change over time. We focus on forecasting and so the parsimony of the model is important for good performance. A prior is developed which allows the shrinkage of the regression coefficients to suitably change over time and an efficient Markov chain Monte Carlo method for posterior inference is described. The new method is applied to two forecasting problems in econometrics: equity premium prediction and inflation forecasting. The results show that this method outperforms current competing Bayesian methods.     

Bayesian Fuzzy Regression Discontinuity Analysis and Returns to Compulsory Schooling

This paper is concerned with the use of a Bayesian approach to fuzzy regression discontinuity (RD) designs for understanding the returns to education. The discussion is motivated by the change in government policy in the UK in April of 1947, when the minimum school leaving age was raised from 14 to 15—a change that had a discontinuous impact on the probability of leaving school at age 14 for cohorts who turned 14 around the time of the policy change. We develop a Bayesian fuzzy RD framework that allows us to take advantage of this discontinuity to calculate the effect of an additional year of education on subsequent log earnings for the (latent) class of subjects that complied with the policy change. We illustrate this approach with a new dataset composed from the UK General Household Surveys. Copyright © 2015 John Wiley & Sons, Ltd.     

FFORMPP: Feature-based forecast model performance prediction

This paper introduces a novel meta-learning algorithm for time series forecast model performance prediction. We model the forecast error as a function of time series features calculated from historical time series with an efficient Bayesian multivariate surface regression approach. The minimum predicted forecast error is then used to identify an individual model or a combination of models to produce the final forecasts. It is well known that the performance of most meta-learning models depends on the representativeness of the reference dataset used for training. In such circumstances, we augment the reference dataset with a feature-based time series simulation approach, namely GRATIS, to generate a rich and representative time series collection. The proposed framework is tested using the M4 competition data and is compared against commonly used forecasting approaches. Our approach provides comparable performance to other model selection and combination approaches but at a lower computational cost and a higher degree of interpretability, which is important for supporting decisions. We also provide useful insights regarding which forecasting models are expected to work better for particular types of time series, the intrinsic mechanisms of the meta-learners, and how the forecasting performance is affected by various factors.     

混合地理加权回归模型中的空间相关性检验和参数估计研究

对混合地理加权回归模型,提出新的空间相关性检验统计量,利用三阶矩?2逼近方法导出了其检验p-值的近似计算公式,模拟结果显示该检验统计量在检测空间相关性方面具有满意的功效。为了处理数据中可能同时存在的空间相关性和空间异质性,引入一类新的混合地理加权空间滞后回归模型,模拟结果表明该估计方法具有较高的可靠性和稳健性,与全局空间滞后回归模型比较,混合地理加权空间滞后回归模型在处理空间异质性方面具有更优良的表现。     

A fuzzy bi-level method for modeling age-specific migration

Modeling and forecasting international migration are significant research areas since migration forecasts are vital in decision making and policy design regarding economy, security, society, and resource allocation. The methods for modeling and forecasting migration rely on strict subjective or statistical assumptions which may not always be met. In addition, lack of a universally accepted definition of the term “migrant” and the ambiguities in data due to recording and collection systems result in inconsistencies and vagueness in migration modeling. Considering these, in this paper, a fuzzy bi-level age-specific migration modeling method is proposed. The bi-level structure embedded in the model makes use of the well-known Lee-Carter method as well as fuzzy regression, singular value decomposition technique, and hierarchical clustering to reflect the general characteristics of the country of concern together with the distinct emigration and immigration behaviors of the age groups. Bayesian time series models are fitted to the time-variant fuzzy parameters obtained through the proposed method to forecast future migration values. The proposed method is applied on female and male age-specific emigration and immigration counts of Finland for 1990–2010 period and Germany for 1995–2012 period, and the future values are forecasted for 2011–2025 and 2013–2025 respectively. The method is compared with an existing Bayesian approach and the numerical findings display that the proposed fuzzy method is superior to the existing one in modeling and forecasting age-specific migration values within significantly narrower prediction intervals.     

Bayesian estimation of proportion and sensitivity level in randomized response procedures

A Bayesian approach to the joint estimation of population proportion and sensitivity level of a stigmatizing attribute is proposed by adopting a two-stage randomized response procedure. In the first stage the direct question method is carried out for each respondent, while in the second stage the randomization is exclusively carried out for those individuals declaring their membership in the non-sensitive group. The randomization is implemented on the basis of Franklin’s procedure. The proposed Bayesian method avoids the drawbacks usually connected with the use of maximum-likelihood or moment estimation.     

Bayesian portfolio selection using a multifactor model

This article develops a new portfolio selection method using Bayesian theory. The proposed method accounts for the uncertainties in estimation parameters and the model specification itself, both of which are ignored by the standard mean-variance method. The critical issue in constructing an appropriate predictive distribution for asset returns is evaluating the goodness of individual factors and models. This problem is investigated from a statistical point of view; we propose using the Bayesian predictive information criterion. Two Bayesian methods and the standard mean-variance method are compared through Monte Carlo simulations and in a real financial data set. The Bayesian methods perform very well compared to the standard mean-variance method.     

Penalized Regression with Ordinal Predictors

Ordered categorial predictors are a common case in regression modelling. In contrast to the case of ordinal response variables, ordinal predictors have been largely neglected in the literature. In this paper, existing methods are reviewed and the use of penalized regression techniques is proposed. Based on dummy coding two types of penalization are explicitly developed; the first imposes a difference penalty, the second is a ridge type refitting procedure. Also a Bayesian motivation is provided. The concept is generalized to the case of non-normal outcomes within the framework of generalized linear models by applying penalized likelihood estimation. Simulation studies and real world data serve for illustration and to compare the approaches to methods often seen in practice, namely simple linear regression on the group labels and pure dummy coding. Especially the proposed difference penalty turns out to be highly competitive.     

Differences Between Classical and Bayesian Estimates for Mixed Logit Models: A Replication Study

The mixed logit model is widely used in applied econometrics. Researchers typically rely on the free choice between the classical and Bayesian estimation approach. However, empirical evidence of the similarity of their parameter estimates is sparse. The presumed similarity is mainly based on one empirical study that analyzes a single dataset (Huber J, Train KE. 2001. On the similarity of classical and Bayesian estimates of individual mean partworths. Marketing Letters 12 (3): 259–269). Our replication study offers a generalization of their results by comparing classical and Bayesian parameter estimates from six additional datasets and specifically for panel versus cross‐sectional data. In general, our results suggest that the two methods provide similar results, with less similarity for cross‐sectional data than for panel data. Copyright © 2016 John Wiley & Sons, Ltd.     

Bayesian clustering with priors on partitions

Traditional clustering algorithms are deterministic in the sense that a given dataset always leads to the same output partition. This article modifies traditional clustering algorithms whereby data are associated with a probability model, and clustering is carried out on the stochastic model parameters rather than the data. This is done in a principled way using a Bayesian approach which allows the assignment of posterior probabilities to output partitions. In addition, the approach incorporates prior knowledge of the output partitions using Bayesian melding. The methodology is applied to two substantive problems: (i) a question of stylometry involving a simulated dataset and (ii) the assessment of potential champions of the 2010 FIFA World Cup.     

Bayesian modeling of joint and conditional distributions

In this paper, we study a Bayesian approach to flexible modeling of conditional distributions. The approach uses a flexible model for the joint distribution of the dependent and independent variables and then extracts the conditional distributions of interest from the estimated joint distribution. We use a finite mixture of multivariate normals (FMMN) to estimate the joint distribution. The conditional distributions can then be assessed analytically or through simulations. The discrete variables are handled through the use of latent variables. The estimation procedure employs an MCMC algorithm. We provide a characterization of the Kullback–Leibler closure of FMMN and show that the joint and conditional predictive densities implied by the FMMN model are consistent estimators for a large class of data generating processes with continuous and discrete observables. The method can be used as a robust regression model with discrete and continuous dependent and independent variables and as a Bayesian alternative to semi- and non-parametric models such as quantile and kernel regression. In experiments, the method compares favorably with classical nonparametric and alternative Bayesian methods.     

Acceleration of the stochastic search variable selection via componentwise Gibbs sampling

The stochastic search variable selection proposed by George and McCulloch (J Am Stat Assoc 88:881–889, 1993) is one of the most popular variable selection methods for linear regression models. Many efforts have been proposed in the literature to improve its computational efficiency. However, most of these efforts change its original Bayesian formulation, thus the comparisons are not fair. This work focuses on how to improve the computational efficiency of the stochastic search variable selection, but remains its original Bayesian formulation unchanged. The improvement is achieved by developing a new Gibbs sampling scheme different from that of George and McCulloch (J Am Stat Assoc 88:881–889, 1993). A remarkable feature of the proposed Gibbs sampling scheme is that, it samples the regression coefficients from their posterior distributions in a componentwise manner, so that the expensive computation of the inverse of the information matrix, which is involved in the algorithm of George and McCulloch (J Am Stat Assoc 88:881–889, 1993), can be avoided. Moreover, since the original Bayesian formulation remains unchanged, the stochastic search variable selection using the proposed Gibbs sampling scheme shall be as efficient as that of George and McCulloch (J Am Stat Assoc 88:881–889, 1993) in terms of assigning large probabilities to those promising models. Some numerical results support these findings.     

Semi-parametric dynamic asymmetric Laplace models for tail risk forecasting,incorporating realized measures

This paper extends the joint Value-at-Risk (VaR) and expected shortfall (ES) quantile regression model of Taylor (2019), by incorporating a realized measure to drive the tail risk dynamics, as a potentially more efficient driver than daily returns. Furthermore, we propose and test a new model for the dynamics of the ES component. Both a maximum likelihood and an adaptive Bayesian Markov chain Monte Carlo method are employed for estimation, the properties of which are compared in a simulation study. The results favour the Bayesian approach, which is employed subsequently in a forecasting study of seven financial market indices. The proposed models are compared to a range of parametric, non-parametric and semi-parametric competitors, including GARCH, realized GARCH, the extreme value theory method and the joint VaR and ES models of Taylor (2019), in terms of the accuracy of one-day-ahead VaR and ES forecasts, over a long forecast sample period that includes the global financial crisis in 2007–2008. The results are favorable for the proposed models incorporating a realized measure, especially when employing the sub-sampled realized variance and the sub-sampled realized range.     

贝叶斯非线性分层模型在多元索赔准备金评估中的应用

本文将贝叶斯非线性分层模型应用于基于不同业务线的多元索赔准备金评估中,设计了一种合适的模型结构,将非线性分层模型与贝叶斯方法结合起来,应用WinBUGS软件对精算实务中经典流量三角形数据进行建模分析,并使用MCMC方法得到了索赔准备金完整的预测分布。这种方法扩展并超越了已有多元评估方法中最佳估计和预测均方误差估计的研究范畴。在贝叶斯框架下结合后验分布实施推断对非寿险公司偿付能力监管和行业决策具有重要作用。     

Bayesian quantile regression methods

This paper is a study of the application of Bayesian exponentially tilted empirical likelihood to inference about quantile regressions. In the case of simple quantiles we show the exact form for the likelihood implied by this method and compare it with the Bayesian bootstrap and with Jeffreys' method. For regression quantiles we derive the asymptotic form of the posterior density. We also examine Markov chain Monte Carlo simulations with a proposal density formed from an overdispersed version of the limiting normal density. We show that the algorithm works well even in models with an endogenous regressor when the instruments are not too weak. Copyright © 2009 John Wiley & Sons, Ltd.     

Analysis of binary data with the possibility of wrong ascertainment

Most genetic studies recruit high‐risk families, and the discoveries are based on non‐random selected groups. We must consider the consequences of this ascertainment process to apply the results of genetic research to the general population. In addition, in epidemiological studies, binary responses are often misclassified. We proposed a binary logistic regression model that provides a novel and flexible way to correct for misclassification in binary responses, taking into account the ascertainment issues. A hierarchical Bayesian analysis using Markov chain Monte Carlo method has been carried out to investigate the effect of covariates on disease status. The focus of this paper is to study the effect of classification errors and non‐random ascertainment on the estimates of the model parameters. An extensive simulation study indicated that the proposed model results in substantial improvement of the estimates. Two data sets have been revisited to illustrate the methodology.     

A fully Bayesian tracking algorithm for mitigating disparate prediction misclassification

We develop a fully Bayesian tracking algorithm with the purpose of providing classification prediction results that are unbiased when applied uniformly to individuals with differing sensitive variable values, e.g., of different races, sexes, etc. Here, we consider bias in the form of group-level differences in false prediction rates between the different sensitive variable groups. Given that the method is fully Bayesian, it is well suited for situations where group parameters or regression coefficients are dynamic quantities. We illustrate our method, in comparison to others, on simulated datasets and two real-world datasets.