A sequential sparse polynomial chaos expansion using Bayesian regression for geotechnical reliability estimations

Polynomial chaos expansions (PCEs) have been widely employed to estimate failure probabilities in geotechnical engineering. However, PCEs suffer from two deficiencies: (a) PCE coefficients are solved by the least-square minimization method which easily causes overfitting issues; (b) building a high order PCE is often computationally expensive. In order to overcome the aforementioned drawbacks, the Bayesian regression technique is employed to evaluate PCE coefficients, which not only provides a sparse solution but also avoids overfitting. With the aid of the predictive means and variances given by Bayesian analysis, a learning function is proposed to sequentially select the most informative samples that are critical to build a PCE. This sequential learning scheme can highly enhance the computational efficiency of PCEs. Besides, importance sampling (IS) is incorporated into the sequential learning (SL)-PCEs to deal with geotechnical problems with small failure probabilities. The proposed method of SL-PCE-IS is applied to three illustrative examples, which shows that the improved PCE method is more effective and efficient than the common PCEs method, leading to accurate estimations of small failure probabilities using fewer training samples.     

基于机器学习的稀疏样本下的土壤有机质估算方法

采用GRNN（Generalized Regression Neural Network）和RF（Random Forest）2种机器学习方法构建土壤有机质预测模型,以提高稀疏样本情况下的土壤有机质估算精度。依据北京市大兴区农用地2007年的土壤有机质采样数据,按MMSD准则（Minimization of the Mean of the Shortest Distances）抽稀为8种不同采样密度的样本（分别为2703、1352、676、339、169、85、43、22个样本）,分别采用GRNN、RF和Ordinary kriging对各采样密度下的未知采样点进行预测,采用交叉检验的方式验证各采样密度下未知样点的预测精度。随着采样点密度的下降,样点间的空间自相关性逐渐减弱,半变异函数的拟和精度变差,预测点结果误差增大,预测的置信度降低。当抽稀到43个和22个采样点时,样点间的空间自相关性接近歼灭,半变异函数的决定系数较低且残差较大。普通克里格受到采样点数量和采样密度、样点的空间结构的影响比较明显,其预测精度随采样点数量的下降而下降。在85个采样点及以下时,其预测值与观测值之间没有显著的相关性。GRNN和RF的预测精度受采样密度的影响不大,其预测精度在一个较小的范围内波动,其预测值围绕观测值在一定阈值空间内震荡波动,具有较好的相关性,在85个及以下的采样密度时,预测精度相对普通克里格有较大的提升。普通克里格法不适合在稀疏样本条件下空间插值计算,尤其是在空间自相关性比较弱的情况下。机器学习模型能充分学习土壤间环境信息、样点空间邻近效应信息,兼顾属性相似性和空间自相关,具有更好的稳定性和适应性,不容易受到采样点数量、构型和采样密度等因素的影响,即使在采样点空间自相关性很弱的情况下也能做出稳定预测精度。     

基于Item2Vec负采样优化的专题地图产品个性化推荐方法研究

建立适用于专题地图产品检索的用户偏好推荐模型是提高专题地图质量的有效方式之一,在专题地图产品推荐场景中,存在严重的内容冷启动和评论数据稀疏问题,现有的推荐算法无法为特定类用户推荐不同特征的专题地图产品,导致用户从专题地图中获取偏好信息受到限制。因此,本文构建基于负采样的连续词袋模型和基于Word2Vec的Item2Vec相结合的用户偏好推荐方法,用于专题地图产品推荐。① 计算用户行为日志文件中交互行为数据的隐性评分,以代替专题地图推荐信息流场景中稀疏的用户评论数据;② 基于负采样的连续词袋模型提取目标专题地图的前后地图序列感知特征信息,通过控制正负样本比例为1:2,提升目标专题地图潜在评分的预测精度;③ 通过Item2Vec将带有用户行为特征信息的专题地图映射到向量空间,计算用户对专题地图的相似度矩阵,根据用户偏好程度完成推荐。在构建的专题地图评分实验数据集Thematic CMaps和4个公开验证数据集MovieLens上的测试结果表明:与LFM、Personal Rank、Content Based和SVD 4种传统推荐算法相比,本文所提方法可有效提高潜在评分的预测精度,推荐性能最高达到27.85%;与以霍夫曼采样方式的Item2Vec基础方法和YouTubeNet 2种神经网络推荐算法相比,评分预测精度有一定提高,且推荐性能不断提升,最高达到2.97%和5.78%。以经典算法奇异值分解（SVD）为例,将MovieLens-20M数据集切分后,在数据量不断增大的数据子集中,本文所用方法的评分预测精度和性能均优于SVD方法。     

System Reliability Analysis of an Offshore Jacket Platform

This study investigates strategies for solving the system reliability of large three-dimensional jacket structures.These structural systems normally fail as a result of a series of different components failures.The failure characteristics are investigated under various environmental conditions and direction combinations.Theβ-unzipping technique is adopted to determine critical failure components,and the entire system is simplified as a series-parallel system to approximately evaluate the structural system reliability.However,this approach needs excessive computational effort for searching failure components and failure paths.Based on a trained artificial neural network(ANN),which can be used to approximate the implicit limit-state function of a complicated structure,a new alternative procedure is proposed to improve the efficiency of the system reliability analysis method.The failure probability is calculated through Monte Carlo simulation(MCS)with Latin hypercube sampling(LHS).The features and applicability of the above procedure are discussed and compared using an example jacket platform located in Chengdao Oilfield,Bohai Sea,China.This study provides a reference for the evaluation of the system reliability of jacket structures.     

GB/T 18316-2008的使用探讨

从抽样检验的基本原理出发,结合对空间数据产品特点的理解,针对空间数据产品的抽样检验,从抽样方案、检验内容、质量评价等方面对现行的抽样检验标准GB/T 18316-2008进行了分析。探讨了该标准存在的缺陷对客观、科学、公正评价空间数据产品质量的影响以及解决方法,对GB/T18316-2008的修订有一定借鉴意义。     

地理要素的分布式空间近似查询处理方法

针对传统空间查询无法满足地理数据交互式可视化对处理时间要求的问题,以窗口查询为例,提出了一种空间近似查询处理方法。该方法包括预处理和查询两步:在预处理阶段,利用分布化的线简化算法对空间对象进行顾及误差的预处理采样,将采样过程及误差值用树型结构保存;在查询阶段,以豪斯多夫距离定义数据可视化的误差,进行误差可知的顶点即时采样与截取,从而实现针对可视化应用的高效的空间近似查询处理。在Hadoop集群上利用77GB的OpenStreetMap数据集进行了实验,证实了本方法的效力与效率。     

Land cover in Norway based on an area frame survey of vegetation types

The Norwegian area frame survey of land cover and outfield land resources (AR18X18), completed in 2014, provided unbiased statistics of land cover in Norway. The article reports the new statistics, discusses implications of the data set, and provides potential value in terms of research, management, and monitoring. A gridded sampling design for 1081 primary statistical units of 0.9 km² at 18?km intervals was implemented in the survey. The plots were mapped in situ, aided by aerial photos, and all areas were coded following a vegetation type system. The results provide new insights into the cover and distribution of vegetation and land cover types. The statistic for mire and wetlands, which previously covered 5.8%, has since been corrected to 8.9%. The survey results can be used for environmental and agricultural management, and the data can be stratified for regional analyses. The survey data can also serve as training data for remote sensing and distribution modelling. Finally, the survey data can be used to calibrate vegetation perturbations in climate change research that focuses on atmospheric–vegetation feedback. The survey documented novel land cover statistics and revealed that the national cover of wetlands had previously been underestimated.     

基于不确定性的大样本集抽样设计

本文研究是针对全球森林覆盖变化检测系统的大训练样本选择的优化问题。样本自动标记系统自动标记了"千万个"代表森林和非森林的像元。为了提高精度和效率,我们需要从千万个样本的大样本集中选择对训练分类器有益的样本。本文提出了两种方法:第一种方法是传统的分层等距离采样;第二种方法是基于不确定性的样本优化,通过研究相同样本间的相似性度量指标、样本不确定性度量指标及样本特征空间分布,来实现基于不确定性的训练样本选择策略。实验通过精度评价验证了基于不确定性的采样策略能比传统的分层等距离采样策略获得好的效果。     

利用数字测图软件实现全站仪坐标放样的内外一体化

为了减小目前常用坐标放样方法中手工输入全站仪数据对作业精度和速度的影响,利用CASS、SV300等数字测图软件相应功能提取各设计点的坐标值,把这些设计数据与控制点数据合并成一个数据文件后,将此数据文件通过数据线传输到全站仪内存文件中,现场利用全站仪自带的坐标放样软件,快速准确地进行放样,从而实现了坐标放样的内外一体化。     

不同采样率下四分量钻孔应变观测同震变化特征

利用湟源台四分量钻孔应变观测的分钟值、1 sps、10 sps和100 sps四种不同采样率的观测数据,通过自检分析、同震应变阶分析和频谱分析等方法研究青海玛多M7.4地震同震变化特征。研究结果表明,分钟采样记录的地震波信息缺失严重,用分钟采样数据进行地震波初动、同震变化幅度等研究将会得到信度较低的结论;采样率越高,记录应变地震波信息的能力越强,但100 sps采样和10 sps采样结果相差不大,10 sps采样已能记录到比较全面的应变地震波信息;同震应变阶的变化性质和变化幅度与采样率无关;未来布设四分量钻孔应变仪时,建议应将采样率至少提高至1 sps。