Word Embeddings and Semantic Spaces in Natural Language Processing

One of the critical hurdles, and breakthroughs, in the field of Natural Language Processing (NLP) in the last two decades has been the development of techniques for text representation that solves the so-called curse of dimensionality, a problem which plagues NLP in general given that the feature set for learning starts as a function of the size of the language in question, upwards of hundreds of thousands of terms typically. As such, much of the research and development in NLP in the last two decades has been in finding and optimizing solutions to this problem, to feature selection in NLP effectively. This paper looks at the development of these various techniques, leveraging a variety of statistical methods which rest on linguistic theories that were advanced in the middle of the last century, namely the distributional hypothesis which suggests that words that are found in similar contexts generally have similar meanings. In this survey paper we look at the development of some of the most popular of these techniques from a mathematical as well as data structure perspective, from Latent Semantic Analysis to Vector Space Models to their more modern variants which are typically referred to as word embeddings. In this review of algoriths such as Word2Vec, GloVe, ELMo and BERT, we explore the idea of semantic spaces more generally beyond applicability to NLP.

Higher-Order Statistics and Nonlinear Processes in Space Plasmas

Statistics of order 2 (variance, auto and cross-correlation functions, auto and cross-power spectra) and 3 (skewness, auto and cross-bicorrelation functions, auto and cross-bispectra) are used to analyze the wave-particle interaction in space plasmas. The signals considered here are medium scale electron density irregularities and ELF/ULF electrostatic turbulence. Nonlinearities are mainly observed in the ELF range. They are independently pointed out in time series associated with fluctuations in electronic density and in time series associated with the measurement of one electric field component. Peaks in cross-bicorrelation function and in mutual information clearly show that, in well delimited frequency bands, the wave-particle interactions are nonlinear above a certain level of fluctuations. The way the energy is transferred within the frequencies of density fluctuations is indicated by a bi-spectra analysis.

Range walk and array rotation in space-time adaptive processing:effects and compensations

This paper proposes a unified clutter model incorporating the effects of range walk and array rotation for space-time adaptive processing（STAP） in airborne multi-channel early-warning radar.Based on this clutter model,STAP performance is then analyzed from the perspective of covariance matrix tapering（CMT）.For STAP performance degradation due to array rotation,a determinate compensation method is proposed based on the CMT method.Numerical examples are provided to verify the analysis and the proposed compensation method.

SPACE-TIME ADAPTIVE PROCESSING FOR AIRBORNE RADAR:A CONVENIENT IMPLEMENTATION APPROACH

A convenient implementation approach to space-time adaptive processing for airborne radar has been proposed, which is added by some auxiliary array elements in the area of main-lobe clutter on the basis of 2-D Capon approach . It is of practical use for its small computational load. This approach possesses the ideal performance in the area of main-lobe clutter . In addition, the approach which is added by some auxiliary beams in the area of main-lobe clutter has also been discussed.

Direct data domain approach to space-time adaptive processing

In non-homogeneous environment, traditional space-time adaptive processing doesn＇t effectively suppress interference and detect target, because the secondary data don＇ t exactly reflect the statistical characteristic of the range cell under test. A ravel methodology utilizing the direct data domain approach to space-time adaptive processing （ STAP ） in airbome radar non-homogeneous environments is presented. The deterministic least squares adaptive signal processing technique operates on a ＂snapshot-by-snapshot＂ basis to dethrone the adaptive adaptive weights for nulling interferences and estimating signal of interest （SOI）. Furthermore, this approach eliminates the requirement for estimating the covariance through the data of neighboring range cell, which eliminates calculating the inverse of covariance, and can be implemented to operate in real-time. Simulation results illustrate the efficiency of interference suppression in non-homogeneous environment.

Relationship of public preferences and behavior in residential outdoor spaces using analytic hierarchy process and principal component analysis—a case study of Hangzhou City, China

This study examined public attitudes concerning the value of outdoor spaces which people use daily. Two successive analyses were performed based on data from common residents and college students in the city of Hangzhou, China. First, citizens registered various items constituting desirable values of residential outdoor spaces through a preliminary questionnaire. The result proposed three general attributes (functional, aesthetic and ecological) and ten specific qualities of residential outdoor spaces. An analytic hierarchy process (AHP) was applied to an interview survey in order to clarify the weights among these attributes and qualities. Second, principal factors were extracted from the ten specific qualities with principal component analysis (PCA) for both the common case and the campus case. In addition, the variations of respondents’ groups were classified with cluster analysis (CA) using the results of the PCA. The results of the AHP application found that the public prefers the functional attribute, rather than the aesthetic attribute. The latter is always viewed as the core value of open spaces in the eyes of architects and designers. Fur-thermore, comparisons of ten specific qualities showed that the public prefers the open spaces that can be utilized conveniently and easily for group activities, because such spaces keep an active lifestyle of neighborhood communication, which is also seen to protect human-regarding residential environments. Moreover, different groups of respondents diverge largely in terms of gender, age, behavior and preference.

State Space Model Predictive Control of an Aerothermic Process with Actuators Constraints

This paper investigates State Space Model Predictive Control (SSMPC) of an aerothermic process. It is a pilot scale heating and ventilation system equipped with a heater grid and a centrifugal blower, fully connected through a data acquisition system for real time control. The interaction between the process variables is shown to be challenging for single variable controllers, therefore multi-variable control is worth considering. A multi-variable state space model is obtained from on-line experimental data. The controller design is translated into a Quadratic Programming (QP) problem, in which a cost function subject to actuators linear inequality constraints is minimized. The outcome of the experimental results is that the main control objectives, such as set-point tracking and perturbations rejection under actuators constraints, are well achieved for both controlled variables simultaneously.

PRECONDITIONED METHODS FOR SPACE-TIME ADAPTIVE PROCESSING

This paper introduces the preconditioned methods for Space-Time Adaptive Processing(STAP).Using the Block-Toeplitz-Toeplitz-Block(BTTB)structure of the clutter-plus-noise covari-ance matrix,a Block-Circulant-Circulant-Block(BCCB)preconditioner is constructed.Based on thepreconditioner,a Preconditioned Multistage Wiener Filter(PMWF)which can be implemented by thePreconditioned Conjugate Gradient(PCG)method is proposed.Simulation results show that thePMWF has faster convergence rate and lower processing rank compared with the MWF.

STATISTICAL SPACE-TIME ADAPTIVE PROCESSING ALGORITHM

For the slowly changed environment-range-dependent non-homogeneity, a new statistical space-time adaptive processing algorithm is proposed, which uses the statistical methods, such as Bayes or likelihood criterion to estimate the approximative covariance matrix in the non-homogeneous condition. According to the statistical characteristics of the space-time snapshot data, via defining the aggregate snapshot data and corresponding events, the conditional probability of the space-time snapshot data which is the effective training data is given, then the weighting coefficients are obtained for the weighting method. The theory analysis indicates that the statistical methods of the Bayes and likelihood criterion for covariance matrix estimation are more reasonable than other methods that estimate the covariance matrix with the use of training data except the detected outliers. The last simulations attest that the proposed algorithms can estimate the covariance in the non-homogeneous condition exactly and have favorable characteristics.

A NON－INCREMENTAL TIME－SPACE ALGORITHM FOR NUMERICAL SIMULATION OF FORMING PROCESS

A non-incremental time-space algorithm is proposed for numerical. analysis of forming process with the inclusion of geometrical, material, contact-frictional nonlinearities. Unlike the widely used Newton-Raphson counterpart, the present scheme features an iterative solution procedure on entire time and space domain. Validity and feasibility of foe present scheme are further justiced by the numerical investigation herewith presented.

ISHIKAWA ITERATIVE PROCESS IN UNIFORMLY SMOOTH BANACH SPACES

Let E be a uniformly smooth Banach space, K be a nonempty closed convex subset of E, and suppose: T: K --> K is a continuous Phi-strongly pseudocontractive operator with a bounded range. Using a new analytical method, under general cases, the Ishikawa iterative process {x(n)} converges strongly to the unique fixed point x* of the operator T were proved. The paper generalizes and extends a lot of recent corresponding results.

Nonlinear dynamics and coupling effect of libration and vibration of tethered space robot in deorbiting process

In order to control the growth of space debris,a novel tethered space robot(TSR) was put forward.After capture,the platform,tether,and target constituted a tethered combination system.General nonlinear dynamics of the tethered combination system in the post-capture phase was established with the consideration of the attitudes of two spacecrafts and the quadratic nonlinear elasticity of the tether.The motion law of the tethered combination in the deorbiting process with different disturbances was simulated and discussed on the premise that the platform was only controlled by a constant thrust force.It is known that the four motion freedoms of the tethered combination are coupled with each other in the deorbiting process from the simulation results.A noticeable phenomenon is that the tether longitudinal vibration does not decay to vanish even under the large tether damping with initial attitude disturbances due to the coupling effect.The approximate analytical solutions of the dynamics for a simplified model are obtained through the perturbation method.The condition of the inter resonance phenomenon is the frequency ratio λ_1=2.The case study shows good accordance between the analytical solutions and numerical results,indicating the effectiveness and correctness of approximate analytical solutions.

基于自适应字典校正的稀疏恢复STAP算法

空时自适应处理(Space-Time Adaptive Processing,STAP)技术在时间维(脉冲维)和空间维(阵元维)联合进行信号处理,以实现动目标检测功能。但是,传统STAP技术的计算复杂度非常高,而且在优化处理信号过程中需要大样本的支撑,在实际的工作场景中,杂波环境复杂易变,不易获取足够多的独立同分布样本,因此杂波抑制效果较差。稀疏恢复空时自适应处理(Sparse Recovery Space-Time Adaptive Processing,SR-STAP)算法可以利用很少的训练样本实现杂波抑制,但大多数SR-STAP算法的计算量巨大,运行速度慢,算法实时性不高。此外,SRSTAP算法需要对连续空时二维平面进行离散化处理,将空时二维平面划分为很多细小的网格,由于真实的杂波在空时平面上是连续分布的,同时考虑雷达接收信号中噪声、系统参数误差等因素的影响,真实杂波点与离散化网格点之间一定存在着偏差,会造成网格失配现象,导致SR-STAP算法杂波抑制性能下降。针对此问题,本文提出了基于自适应字典校正的稀疏恢复STAP算法。该算法首先通过子空间投影法筛选出与杂波最相关的原子;然后围绕选定原子由粗到细进行自适应局部网格划分,按照局部网格迭代选优准则,不断调整选择局域内的最优原子,直到满足迭代终止条件,以匹配真实的杂波点;最后利用选定的最优原子对应的空时导向矢量构造杂波子空间,更新噪声子空间上与杂波子空间正交的投影矩阵得到STAP权值。仿真实验表明,所提算法与传统SR-STAP算法相比,具有更高的稀疏恢复精度,更快的运行速度,改善了STAP性能。

基于QbD理念的散偏汤提取工艺设计空间预测

目的:预测散偏汤提取工艺的设计空间。方法:基于质量源于设计和设计空间理念,采用鱼骨图结合失效模型与效应分析的方法确定提取次数、提取时间、加水倍量为散偏汤提取工艺的关键工艺参数,以阿魏酸、芥子碱硫氰酸盐、甘草苷、甘草酸的含量及出膏率为关键质量属性,采用Box-Behnken响应面法试验设计拟合数学模型,考察其相互关系,建立散偏汤提取工艺设计空间。结果:依据数学模型建立了散偏汤提取工艺设计空间,在此空间内取点验证,发现单个处方阿魏酸、芥子碱硫氰酸盐、甘草苷、甘草酸的质量分数分别为0.450‰、0.787‰、0.248‰、0.451‰,出膏率为23.83%,符合关键质量属性的要求。结论:建立的设计空间可为散偏汤放大生产和质量控制提供参考。

信息加工视角下地铁导视系统设计策略研究

目的对地铁导视系统设计现状进行梳理,旨在从信息加工视角出发,探究地铁导视系统设计策略,以提升乘客的寻路体验。方法通过文献和现状调研分析现阶段地铁导视系统设计存在的问题,论述信息加工模型切入的可行性,构建地铁导视信息加工模型。以武汉地铁为例,通过定性和定量对当前导视系统效力进行评估并确定痛点,基于模型从感知处理、记忆储存、决策执行三个层面对地铁导视系统痛点进行分类分析。结论构建了地铁导视信息加工模型,提出了以感知为基础的表现层设计、以记忆为依托的结构层设计、以决策为主线的行为层设计三条策略。从信息加工视角审视地铁导视系统设计,不仅可以提升乘客的寻路体验,还能有效提高地铁系统的整体运行效率。

Iterative Processes with Errors for Generalized Set-Valued φ-Hemi-contractive Mapping in Banach Spaces

A new conception of generalized set-valued Ф-hemi-contractive mapping in Banach spaces is presented. Some strong convergence theorems of Ishikawa and Mann iterative approximation with errors is proved. The results in this paper improve and extend the earlier results.

空频抗干扰对GNSS载波相位测量的影响

对于全球导航卫星系统(GNSS)的抗干扰,最有效的是采用自适应天线阵技术,但这种抗干扰处理会引入与信号入射方向相关的载波相位测量偏差,限制了其在高精确度测量领域的应用。为减小抗干扰处理的偏差,引入测量偏差及其分析方法,建立仿真模型,并搭建软件接收机进行实验。结果表明,在不增加额外约束的条件下,采用空频最小方差无畸变响应算法实现空频抗干扰处理器,不会引入载波相位测量偏差,非常适用于对抗干扰性能和测量精确度均有苛刻要求的场合。

天津红色文化遗产空间价值评估研究

红色文化遗产空间是近年来城市遗产研究中的一个重要方向,对于延续城市历史文脉、坚定文化自信具有深远意义。以天津为例,通过收集、整理天津市域范围内159处红色文化遗产空间的数字信息,运用层次分析法和GIS定量化地评估红色文化遗产空间的开发利用价值,并将评估结果进行可视化表达。研究表明天津市红色文化遗产空间具有类型多元和保护状况较好的特征,但未形成有序的空间序列,在群众认知中价值差异较大。研究成果可为天津市红色文化遗产空间未来的保护与利用提供科学依据。

多模式对四川盆地强降水过程的预报性能检验

为进一步认识当前数值预报模式的预报能力,选取2018—2020年发生在四川盆地的47次强降水过程进行分型,再基于多源降水融合产品和地面观测资料,通过TS评分、时空滑动等方法对欧洲中期天气预报中心(European Centre for Medium-Range Weather Forecasts,ECMWF)数值预报模式、国家气象中心区域中尺度数值预报模式(China Meteorological Administration Mesoscale Model,CMA_MESO)和西南区域数值预报系统(Southwest Center WRF ADAS Real-time Modeling System,SWC_WARMS)在强降水过程范围、强度、极值、时间和位移偏差等方面的预报能力进行检验评估。结果表明,各模式08:00(北京时,下同)预报优于20:00预报,ECMWF对中雨和大雨预报更优,SWC_WARMS的暴雨量级评分更高。各模式对中雨的预报范围普遍较实况偏大,随着降水量级增大,逐渐转为低估,其中SWC_WARMS更接近实况。对于降水强度,ECMWF和CMA_MESO的平均降水量和极值普遍较实况偏小,SWC_WARMS更接近实况。3种模式时间偏差不明显,仅个别起报时次有-6~3 h的时间偏差;ECMWF的位移偏差最小,纬向上ECMWF和SWC_WARMS以偏北为主,经向上ECMWF以偏西为主,CMA_MESO和SWC_WARMS以偏东为主。

基于KA-SRCN-pSTAP的低空风切变风速估计方法

针对由于独立同分布(IID)样本严重不足,导致极化空时自适应(pSTAP)处理性能下降,进而导致低空风切变风速估计不准确的问题,本文提出了一种基于知识辅助的稀疏表示杂波零陷极化空时自适应(KA-SRCN-pSTAP)的低空风切变风速估计方法。该方法首先利用杂波脊的先验知识辅助构造极化空时稀疏字典,然后利用极化空时稀疏字典,通过SRCN算法挑选原子并对到杂波线性子空间补空间上的投影矩阵进行估计,从而得到pSTAP权矢量,最后构造pSTAP滤波器对地杂波进行抑制,准确估计低空风切变风速。该方法仅使用少量IID样本,将SRCN算法与极化-空时域相结合,完成对风切变风速的有效估计。仿真实验结果证明该方法可以有效实现少样本情况下的风速准确估计。