Global Robustness with Respect to the Loss Function and the Prior

We propose a class [I,S] of loss functions for modeling the imprecise preferences of the decision maker in Bayesian Decision Theory. This class is built upon two extreme loss functions I and S which reflect the limited information about the loss function. We give an approximation of the set of Bayes actions for every loss function in [I,S] and every prior in a mixture class; if the decision space is a subset of , we obtain the exact set.     

A Bayesian Benefit-Risk Model Applied to the South Florida Building Code

A Bayesian compound Poisson benefit-risk model is described in this paper, and used to evaluate recent revisions to the South Florida Building Code (SFBC). The model accounts for natural variability in hurricane frequency and severity, and uncertainty in the effectiveness of the revised code. Ranges of residential growth rate, code effectiveness, construction cost increase, and planning period length are assumed, to show the ranges of cost-to-performance ratio within which the code will make sense economically. The expected cost of residential hurricane damage over 50 years for ten South Florida counties assuming continuation of previous building practices was $93 billion, equivalent to the residential damage of 5.2 Andrews. Assuming a reduction in the growth of damageable housing in South Florida from 5.5% to 2% as a result of code revision, estimated damages under the new code were $45 billion. At a per-house construction cost increase of 5%, the probability of at least recovering the estimated $40 billion cost of the specified wind-resistant construction was estimated to be 47%. Expected return on investment was estimated at $7 billion over 50 years. The expected return lies between a $44 billion loss and a $47 billion gain, when growth in damageable housing is allowed to range from 1% to 4% and construction cost increases are assumed to lie between 3% and 8%. Actual monetary return for a 5% cost increase and 2% growth in damageable housing ranges from a $20 billion loss to a $100 billion gain with 95% probability, as a result of weather variability alone. Results support SFBC revisions on solely economic grounds, a conclusion strengthened considerably in light of potentially avoided deaths and hurricane traumas. The model represents one approach to evaluating economic aspects of the sustainability of new technological measures on the basis of available information.     

Challenges to the Acceptance of Probabilistic Risk Analysis

This paper discusses a number of the key challenges to the acceptance and application of probabilistic risk analysis (PRA). Those challenges include: (a) the extensive reliance on subjective judgment in PRA, requiring the development of guidance for the use of PRA in risk-informed regulation, and possibly the development of robust or reference prior distributions to minimize the reliance on judgment; and (b) the treatment of human performance in PRA, including not only human error per se but also management and organizational factors more broadly. All of these areas are seen as presenting interesting research challenges at the interface between engineering and other disciplines.     

"假象"背后的均衡及信息的代价研究

通过库诺特模型可以说明,“假象”的存在将改变均衡结果:如果信息获取是无成本的,那么“假象”将使信息占优一方的均衡利润增加,其对手的均衡利润可能减少,也可能增加;如果获取对手的信息是有代价的,根据我们给出的信息投资的边界条件,当获取信息的代价大于这一边界条件时,获取对手信息是有必要的,反之,应该放弃对信息的投资。     

Decision Making Under Risk: A Comparison of Bayesian and Fuzzy Set Methods

A classical decision problem is considered where a decision maker is to choose one of a number of actions each offering different consequences. The outcome from a choice of action is uncertain because it depends on the existing state of Nature. Also, the outcome, once an action and state of Nature are specified, may be a vector or a random vector. The decision maker employs both Bayesian methods and fuzzy set techniques to handle the uncertainties. The decision maker is also allowed to use multiple, possibly conflicting, goals in order to determine his best strategy. The Bayesian method produces a set of undominated strategies to choose from, whereas the fuzzy set technique usually produces a unique optimal strategy.     

Fusion of Intelligence Information: A Bayesian Approach

The attack that occurred on September 11, 2001 was, in the end, the result of a failure to detect and prevent the terrorist operations that hit the United States. The U.S. government thus faces at this time the daunting tasks of first, drastically increasing its ability to obtain and interpret different types of signals of impending terrorist attacks with sufficient lead time and accuracy, and second, improving its ability to react effectively. One of the main challenges is the fusion of information, from different sources (U.S. or foreign), and of different types (electronic signals, human intelligence. etc.). Fusion thus involves two very distinct and separate issues: communications, i.e., ensuring that the different U.S. and foreign intelligence agencies communicate all relevant and accurate information in a timely fashion and, perhaps more difficult, merging the content of signals, some "sharp" and some "fuzzy," some dependent and some independent into useful information. The focus of this article is on the latter issue, and on the use of the results. In this article, I present a classic probabilistic Bayesian model sometimes used in engineering risk analysis, which can be helpful in the fusion of information because it allows computation of the posterior probability of an event given its prior probability (before the signal is observed) and the quality of the signal characterized by the probabilities of false positive and false negative. Experience suggests that the nature of these errors has been sometimes misunderstood; therefore, I discuss the validity of several possible definitions.     

Use of Markov Chain Monte Carlo Analysis with a Physiologically-Based Pharmacokinetic Model of Methylmercury to Estimate Exposures in U.S. Women of Childbearing Age

A Bayesian approach, implemented using Markov Chain Monte Carlo (MCMC) analysis, was applied with a physiologically‐based pharmacokinetic (PBPK) model of methylmercury (MeHg) to evaluate the variability of MeHg exposure in women of childbearing age in the U.S. population. The analysis made use of the newly available National Health and Nutrition Survey (NHANES) blood and hair mercury concentration data for women of age 16–49 years (sample size, 1,582). Bayesian analysis was performed to estimate the population variability in MeHg exposure (daily ingestion rate) implied by the variation in blood and hair concentrations of mercury in the NHANES database. The measured variability in the NHANES blood and hair data represents the result of a process that includes interindividual variation in exposure to MeHg and interindividual variation in the pharmacokinetics (distribution, clearance) of MeHg. The PBPK model includes a number of pharmacokinetic parameters (e.g., tissue volumes, partition coefficients, rate constants for metabolism and elimination) that can vary from individual to individual within the subpopulation of interest. Using MCMC analysis, it was possible to combine prior distributions of the PBPK model parameters with the NHANES blood and hair data, as well as with kinetic data from controlled human exposures to MeHg, to derive posterior distributions that refine the estimates of both the population exposure distribution and the pharmacokinetic parameters. In general, based on the populations surveyed by NHANES, the results of the MCMC analysis indicate that a small fraction, less than 1%, of the U.S. population of women of childbearing age may have mercury exposures greater than the EPA RfD for MeHg of 0.1 μg/kgg/day, and that there are few, if any, exposures greater than the ATSDR MRL of 0.3 μgg/kgg/day. The analysis also indicates that typical exposures may be greater than previously estimated from food consumption surveys, but that the variability in exposure within the population of U.S. women of childbearing age may be less than previously assumed.     

时间贝叶斯网络及其概率推理

针对贝叶斯网络应用中出现的循环和动态问题,研究贝叶斯网络的时间扩展.给出了时间贝叶斯网络的定义;探讨了时间贝叶斯网络中环的存在合理性判断问题;给出了时间贝叶斯网络的概率推理算法,应用示例说明了方法的可行性.     

A joint modelling approach for clustered recurrent events and death events

In dental implant research studies, events such as implant complications including pain or infection may be observed recurrently before failure events, i.e. the death of implants. It is natural to assume that recurrent events and failure events are correlated to each other, since they happen on the same implant (subject) and complication times have strong effects on the implant survival time. On the other hand, each patient may have more than one implant. Therefore these recurrent events or failure events are clustered since implant complication times or failure times within the same patient (cluster) are likely to be correlated. The overall implant survival times and recurrent complication times are both interesting to us. In this paper, a joint modelling approach is proposed for modelling complication events and dental implant survival times simultaneously. The proposed method uses a frailty process to model the correlation within cluster and the correlation within subjects. We use Bayesian methods to obtain estimates of the parameters. Performance of the joint models are shown via simulation studies and data analysis.     

Estimation under modified Weibull distribution based on right censored generalized order statistics

In this paper, the maximum likelihood (ML) and Bayes, by using Markov chain Monte Carlo (MCMC), methods are considered to estimate the parameters of three-parameter modified Weibull distribution (MWD(β, τ, λ)) based on a right censored sample of generalized order statistics (gos). Simulation experiments are conducted to demonstrate the efficiency of the proposed methods. Some comparisons are carried out between the ML and Bayes methods by computing the mean squared errors (MSEs), Akaike's information criteria (AIC) and Bayesian information criteria (BIC) of the estimates to illustrate the paper. Three real data sets from Weibull(α, β) distribution are introduced and analyzed using the MWD(β, τ, λ) and also using the Weibull(α, β) distribution. A comparison is carried out between the mentioned models based on the corresponding Kolmogorov–Smirnov (K–S) test statistic, {AIC and BIC} to emphasize that the MWD(β, τ, λ) fits the data better than the other distribution. All parameters are estimated based on type-II censored sample, censored upper record values and progressively type-II censored sample which are generated from the real data sets.