Estimating posterior inference quality of the relational infinite latent feature model for overlapping community detection

Overlapping community detection has become a very hot research topic in recent decades, and a plethora of methods have been proposed. But, a common challenge in many existing overlapping community detection approaches is that the number of communities K must be predefinedmanually. We propose a flexible nonparametric Bayesian generative model for count-value networks, which can allow K to increase as more and more data are encountered instead of to be fixed in advance. The Indian buffet process was used to model the community assignment matrix Z, and an uncollapsed Gibbs sampler has been derived.However, as the community assignment matrix Z is a structured multi-variable parameter, how to summarize the posterior inference results and estimate the inference quality about Z, is still a considerable challenge in the literature. In this paper, a graph convolutional neural network based graph classifier was utilized to help to summarize the results and to estimate the inference quality about Z. We conduct extensive experiments on synthetic data and real data, and find that empirically, the traditional posterior summarization strategy is reliable.     

新型机械搅拌种分槽物料均匀度检测方法初探

本文分析了影响氧化铝生产新型机械搅拌种分槽均匀度检测准确度的主要因素，根据新型机械搅拌种分槽内物料流动型态，探索了均匀度的检测方法，同时采用自行设计的取样器对某厂种分槽均匀度进行了检测。     

Sampling Rate Evaluation Of NO2 Badge: (I) In Indoor Environments

The sampling rate of a nitrogen dioxide (NO²) passive sampling badge was evaluated in indoor environments including an unoccupied research house, residential houses, and an office. Measurements from the NO² badges were compared with those of a chemiluminescent analyzer the EPA reference method, by placing them near to the sample inlet of the chemiluminescent analyzer In this study, we used a new sampling rate for the NO² badge placed in indoor environments (an overall mass transfer coefficient of 0.10 cm/s) smaller than the rate previously reported for the badge when used outdoors. The new rate provides more accurate measurements of NO, concentrations in indoor environments. Indoor NO² concentrations were also measured with the NO² badges exposed to a constant wind velocity provided by a wind tunnel. Since the measurements of the badge with a constant wind velocity agreed well with the reference method, the badges could be assumed to be a secondary reference measurement. With the badges used as the secondary reference measurement, we developed a portable wind tunnel to evaluate a personal exposure measurement by the badge. The results are presented in Environment International (Lee et al., 1993). Precision of the badge measurements was as good as an intraclass correlation coefficient of 0.9779. It was determined that placement of the badge should be at least 10 cm out from an indoor wall surface to avoid undersampling due to NO² gradients near the surface.     

利用衰减全反射傅立叶变换红外光谱方法对乙二醇水溶液定量分析的研究

利用衰减全反射-傅立叶变换红外光谱法(ATR-FTIR)和OMNI采样器附件对乙二醇水溶液进行定量分析的研究。吸光度测量值有较宽的线性范围,重复性好。这一分析方法方便快捷、分析结果准确。     

取样扰动对土的工程性质指标影响的试验研究

本文报道了取样扰动对土的工程性质指标影响的一批研究成果。采用100mm和75mm两种直径的取土器取样作对比试验以研究取土器扰动的影响。试验证明取土器直径对试样的孔隙比、压缩系数、超固结比和天然强度等指标都有明显的影响。采用两种预固结处理的方法以恢复原生应力状态,用不固结不排水试验比较了应力状态对天然强度的影响。对取土过程中土体应变的分布及其对土样扰动的影响用有限元方法进行了模拟。     

Monitoring of volatile organic compounds in non-residential indoor environments

A weekly monitoring campaign of volatile organic compounds (VOC), with single sampling of 24 h, was carried out in non-residential indoor environments such as libraries, pharmacies, offices, gymnasiums, etc., in order to evaluate the VOC concentrations to which people are exposed. Moreover, an outdoor sample was coupled to each indoor site to point out the influence of indoor sources. They were sampled with Radiello diffusive samplers for thermal desorption and analyzed by GC-MS. As already described in other papers, the VOC levels of most of the indoor sites were higher than that observed in the corresponding outdoor sites. For example, some sites showed a level of pollution that is ten times higher than their corresponding outdoor site. The monitored environments that had higher concentrations of the investigated VOC were the pharmacies, a newspaper stand, a copy center, and the coffee shops. Analysis of the weekly average concentrations of each pollutant and the use of literature allowed pointing out some site-specific characteristics that singled out possible sources of VOC. These results were verified analyzing the indoor-outdoor ratio (I/O) too. Newspaper stands were characterized by very high concentrations of toluene and pharmacies were characterized by high concentrations of aromatic compounds. PRACTICAL IMPLICATIONS: Indoor air pollution caused by volatile organic compounds (VOC) might affect human health at home as well as in public and commercial buildings. The main VOC sources in indoor environments are human activities, personal care products, smoking, house cleaning products, building products, and outside pollution. To preserve human health it is necessary to evaluate the average concentrations of VOC to which people are exposed and to identify the main sources of indoor pollution by means of suitable indoor monitoring campaigns in several environments. These investigations allow pointing out the characteristic critical situations of some indoor environments or some other types of environments.     

原状土的无侧限抗压强度试验研究

针对连云港地区滨海相软土的工程特性，通过大量的试验，分析了原状土的无侧限抗压强度的主要影响因素，以确保提高无侧限抗压强度试验的精度。     

放射性气溶胶冲击式采样器的设计与研究

为了提高冲击式采样器的捕集效率,以放射性气溶胶采样器为研究对象,利用斯托克原理计算冲击板的各个参数,并采用CFD软件对采样器喷嘴口处的空气场进行流场仿真。结果表明:减少采样器冲击板的冲击长度S可以缩小喷嘴处流场的负压区域,进而提高采样器捕集效率。     

FFS-A型地下水定深取样器

地下水污染调查需要采集地下水样品进行分析以确定地下水污染程度。为了采集具有代表性的地下水样品,自行研制了FFS-A型地下水定深取样器。对该取样器工作原理、制作材料及野外试验等进行了介绍。     

Multiscale Bayesian texture segmentation using neural networks and Markov random fields

This paper presents a wavelet-based texture segmentation method using multilayer perceptron (MLP) networks and Markov random fields (MRF) in a multi-scale Bayesian framework. Inputs and outputs of MLP networks are constructed to estimate a posterior probability. The multi-scale features produced by multi-level wavelet decompositions of textured images are classified at each scale by maximum a posterior (MAP) classification and the posterior probabilities from MLP networks. An MRF model is used in order to model the prior distribution of each texture class, and a factor, which fuses the classification information through scales and acts as a guide for the labeling decision, is incorporated into the MAP classification of each scale. By fusing the multi-scale MAP classifications sequentially from coarse to fine scales, our proposed method gets the final and improved segmentation result at the finest scale. In this fusion process, the MRF model serves as the smoothness constraint and the Gibbs sampler acts as the MAP classifier. Our texture segmentation method was applied to segmentation of gray-level textured images. The proposed segmentation method shows better performance than texture segmentation using the hidden Markov trees (HMT) model and the HMTseg algorithm, which is a multi-scale Bayesian image segmentation algorithm.