基于贝叶斯网络的电力工程项目延迟风险研究

将电力工程项目发生延迟的各个风险因素通过BN模型进行了量化,结果显示,天气状况恶劣、项目时间计划过紧、技术工种能力低、缺乏先进设备仪器和供应商不能按时供货这5种风险因素会直接对项目的延迟产生影响;供应商不能按时供货、技术工种能力低、项目计划安排不当3种风险因素对项目的延迟十分敏感.项目经理应该重点控制这些敏感因素.     

基于后验概率制导的B-KNN文本分类方法

针对K最近邻(KNN)方法分类准确率高但分类效率较低的特点,提出基于后验概率制导的贝叶斯K最近邻(B-KNN)方法。利用测试文本的后验概率信息对训练集多路静态搜索树进行剪枝,在被压缩的候选类型空间内查找样本的K个最近邻,从而在保证分类准确率的同时提高KNN方法的效率。实验结果表明,与KNN相比,B-KNN的性能有较大提升,更适用于具有较深层次类型空间的文本分类应用。     

Consumer preferences for food labelling attributes: Comparing direct ranking and best–worst scaling for measurement of attribute importance,preference intensity and attribute dominance

The use of food labelling to convey information about product and process quality and for product differentiation purposes has multiplied. In order to judge the relevance attributed by consumers to such information, valid measurement methods are needed. Such methods are also needed to reveal the probabilistic nature of preference data, so that heterogeneity can be explicitly accounted for. A survey among Swedish residents (n = 506) compared attribute importance rankings for labelling of beef from two formats of best–worst scaling (BWS) with those from standard direct ranking (DR). A choice probability R–index measure was modelled to make the methodological comparison consistent. While earlier studies on labelling of beef were confirmed, BWS and DR did not concur when directly compared, even when using the R–index. BWS improved individual choice predictions compared with DR, and generated a more consistent dominance ordering of attribute importance. These findings suggest that methods used to elicit importance weights or preference ranking may violate transitivity and dominance requirements.     

Quantifying Listeria monocytogenes prevalence and concentration in minced pork meat and estimating performance of three culture media from presence/absence microbiological testing using a deterministic and stochastic approach

Listeria monocytogenes poses a serious threat to public health, and the majority of cases of human listeriosis are associated with contaminated food. Reliable microbiological testing is needed for effective pathogen control by food industry and competent authorities. The aims of this work were to estimate the prevalence and concentration of L. monocytogenes in minced pork meat by the application of a Bayesian modeling approach, and also to determine the performance of three culture media commonly used for detecting L. monocytogenes in foods from a deterministic and stochastic perspective. Samples (n = 100) collected from local markets were tested for L. monocytogenes using in parallel the PALCAM, ALOA and RAPID'L.mono selective media according to ISO 11290-1:1996 and 11290-2:1998 methods. Presence of the pathogen was confirmed by conducting biochemical and molecular tests. Independent experiments (n = 10) for model validation purposes were performed. Performance attributes were calculated from the presence–absence microbiological test results by combining the results obtained from the culture media and confirmative tests. Dirichlet distribution, the multivariate expression of a Beta distribution, was used to analyze the performance data from a stochastic perspective. No L. monocytogenes was enumerated by direct-plating (<10 CFU/g), though the pathogen was detected in 22% of the samples. L. monocytogenes concentration was estimated at 14–17 CFU/kg. Validation showed good agreement between observed and predicted prevalence (error = −2.17%). The results showed that all media were best at ruling in L. monocytogenes presence than ruling it out. Sensitivity and specificity varied depending on the culture-dependent method. None of the culture media was perfect in detecting L. monocytogenes in minced pork meat alone. The use of at least two culture media in parallel enhanced the efficiency of L. monocytogenes detection. Bayesian modeling may reduce the time needed to draw conclusions regarding L. monocytogenes presence and the uncertainty of the results obtained. Furthermore, the problem of observing zero counts may be overcome by applying Bayesian analysis, making the determination of a test performance feasible.     

绿色家具评价方法的初步研究

根据绿色产品的定义和家具产品的特点,建立了绿色家具评价指标体系。针对绿色家具标准不健全与评价指标的信息模糊、不完全的现状,就绿色家具可用评价方法进行了分析研究,为绿色家具的认证提供了合理可行的方案。     

基于面向对象的高分辨率遥感影像建筑物提取研究

遥感影像上的建筑物是一种非常重要的地物.利用高分辨率遥感影像上建筑物的光谱、空间特征及上下文特征等,并通过贝叶斯理论计算对象的归属概率属性,提出了一种面向对象的高分辨率影像建筑物提取方法.实验证明该方法有较好的提取效果.     

基于面向对象概率关系模型的威胁级别评估

针对传统贝叶斯网络对复杂对象建模存在的问题,分析了面向对象的概率关系模型.建立基于面向对象概率关系模型知识库的贝叶斯网络生成算法,建立战场威胁级别评估模型,并实现对威胁级别的评估.仿真结果证明了该方法的有效性,并且具有比贝叶斯网络模型更强的建模能力.     

融合多源信息的维纳过程性能退化产品的可靠性评估

针对维纳过程性能退化产品,提出了一种有效融合先验退化信息、寿命数据以及现场退化数据的可靠性评估方法.首先利用Expectation- Maximization( EM)算法基于先验退化信息和寿命数据信息确定参数的先验分布;其次利用贝叶斯方法对参数进行更新,并在此基础上进行可靠性评估.该方法能根据现场退化数据不断地对可靠性进行更新,实现对产品可靠性的实时评估.最后通过金属化膜电容器可靠性评估实例验证了该方法的适用性和有效性.     

Clarification of passion fruit juice by microfiltration: Analyses of operating parameters,study of membrane fouling and juice quality

In this study, the performance of two membranes were compared – tubular ceramic and hollow fiber poly(imide) – under transmembrane pressure of 0.5 and 1 bar, for the clarification of passion fruit pulp pre-treated by centrifugation and enzymatic treatment at the concentrations of 150 and 300 ppm. Nutritional and sensorial qualities of the clarified juice obtained were evaluated. Thus, it was possible to observe that the most adequate condition for the clarification of passion fruit pulp was with enzymatic treatment at 150 ppm and its posterior microfiltration at the ceramic tubular membrane of 0.3 μm with transmembrane pressure of 0.5 bar. The fouling mechanism was identified by estimation of model parameters according to a nonlinear regression by Bayesian inference. Analysis of the fouling mechanism results revealed that hollow fiber membrane is controlled by a cake filtration mechanism, and internal pore blocking fouling mechanism controls ceramic tubular membrane.     

Models in a Bayesian framework for inactivation of pectinesterase in a commercial enzyme formulation by pulsed electric fields

The inactivation of pectinesterase (PE) in a commercial enzyme formulation (CEF) under pulsed electric fields (PEF) was studied. Samples of an aqueous solution of the CEF were exposed to exponential waveform pulses for up to 16 ms at electric field intensities ranging from 5 to 24 kV cm^–1. Pulses were delivered in monopolar mode. The observed inactivation of the enzyme was described by several kinetics and regarding the input of electrical energy density (Q) models using Bayesian non-linear regression. Posterior distributions of the characteristic parameters for each kinetic model (based on the Hülsheger or the Weibull equation) and each Q model (based on exponential decay or the Weibull equation) were obtained. Kinetics models based on the Weibull equation showed better goodness and accuracy than the other models.