一种故障诊断专家系统推理机设计分析

文章基于关系数据库技术,针对目前故障诊断专家系统推理匹配冲突、组合爆炸等导致专家系统运行效率低下的缺陷,提出一种构造所有故障现象与原因对应数据库表,应用关系数据库查询搜索技术的推理机制,将大量推理过程提前到专家系统构造或学习阶段完成。专家系统应用时只执行数据查询搜索,大大提高专家系统的执行效率,为相关方面专家系统设计开发提供借鉴。     

Simulation-based inference in dynamic panel probit models: An application to health

This paper considers the determinants of a binary indicator for the existence of functional limitations using seven waves (1991–1997) of the British Household Panel Survey (BHPS). The focal point of our analysis is the contributions of state dependence, heterogeneity and serial correlation in explaining the dynamics of health. To investigate these issues we apply static and dynamic panel probit models with flexible error structures. To estimate the models we use maximum simulated likelihood (MSL) with antithetic acceleration and implement a recently proposed test for the existence of asymptotic bias. The dynamic models show strong positive state dependence. Data from the British Household Panel Survey (BHPS) were supplied by the ESRC Data Archive. Neither the original collectors of the data nor the Archive bear any responsibility for the analysis or interpretations presented here. Funding was provided by the ESRC award no: R000238169-Simulation-based econometric approaches to investigating the interaction of lifestyle and health. The authors would like to thank William Greene for valuable comments on an earlier draft of the paper, Roberto Leon Gonzalez for valuable programming advice, and participants at the iHEA Third International conference, York, 22–25 July 2001 and York Seminars in Health Econometrics (YSHE) for their comments.     

Forecasting an accumulated series based on partial accumulation II: A new Bayesian method for short series with stable seasonal patterns

The problem of forecasting a time series with only a small amount of data is addressed within a Bayesian framework. The quantity to be predicted is the accumulated value of a positive and continuous variable for which partially accumulated data are available. These conditions appear in a natural way in many situations. A simple model is proposed to describe the relationship between the partial and total values of the variable to be forecasted assuming stable seasonality, which is specified in stochastic terms. Analytical results are obtained for both the point forecast and the entire posterior predictive distribution. The proposed technique does not involve approximations. It allows the use of non-informative priors so that implementation may be automatic. The procedure works well when standard methods cannot be applied due to the reduced number of observations. It also improves on previous results published by the authors. Some real examples are included.     

The Piecemeal Approach to Comparative Advertising

When attempting to portray the attractiveness of a brand vis-à-vis its competitors, an ad may make global claims about superiority or specific claims about one or more attributes. A special case of latter is the piecemeal ad in which the advertised brand is compared to a competitor on one attribute, a different competitor on a second attribute, another competitor on a third attribute, and so on. The present research demonstrates the effectiveness of this technique and explores the parameters of its influence. We find that piecemeal messages are persuasive because they make seemingly strong claims in a believable manner. Consumer skepticism appears to arise only when conditions for scrutiny are very favorable.     

Some Aspects Of Neutral To Right Priors

Neutral to right priors are generalizations of Dirichlet process priors that fit in well with right-censored data. These priors are naturally induced by increasing processes with independent increments which, in turn, may be viewed as priors for the cumulative hazard function. This connection together with the Lévy representation of independent increment processes provides a convenient means of studying properties of neutral to right priors.
This article is a review of the theoretical aspects of neutral to right priors and provides a number of new results on their structural properties. Notable among the new results are characterizations of neutral to right priors in terms of the posterior and the cumulative hazard function. We also show that neutral to right priors are of the following nature: Consistency of Bayes' estimates implies consistency of the posterior, and posterior-consistency for complete observations automatically yields posterior-consistency for right-censored data.     

Erratum to ‘Inventory record inaccuracy: Causes and labor effects’

Inventory record inaccuracy (IRI) is a pervasive problem in retailing and causes non-trivial profit loss. In response to retailers’ interest in identifying antecedents and consequences of IRI, we present a study that comprises multiple modeling initiatives. We first develop a dynamic simulation model to compare and contrast impacts of different operational errors in a continuous (Q, R) inventory system through a full-factorial experimental design. While backroom and shelf shrinkage are found to be predominant drivers of IRI, the other three errors related to recording and shelving have negligible impacts on IRI. Next, we empirically assess the relationships between labor availability and IRI using longitudinal data from five stores in a global retail chain. After deriving a robust measure of IRI through Bayesian computation and estimating panel data models, we find strong evidence that full-time labor reduces IRI whereas part-time labor fails to alleviate it. Further, we articulate the reinforcing relationships between labor and IRI by formally assessing the gain of the feedback loop based on our empirical findings and analyzing immediate, intermediate, and long-term impacts of IRI on labor availability. The feedback modeling effort not only integrates findings from simulation and econometric analysis but also structurally explores the impacts of current practices. We conclude by discussing implications of our findings for practitioners and researchers.     

Modelling Overdispersion for Complex Survey Data

The population characteristics observed by selecting a complex sample from a finite identified population are the result of at least two processes: the process which generates the values attached to the units in the finite population, and the process of selecting the sample of units from the population. In this paper we propose that the resulting observations by viewed as the joint realization of both processes. We overcome the inherent difflculty in modelling the joint processes of generation and selection by exploring second moment and other simplifying assumptions. We obtain general expressions for the mean and covariance function of the joint processes and show that several overdispersion models discussed in the literature for the analysis of complex surveys are a direct consequence of our formulation, undere particular sampling schemes and population structures.