Using molecular representations in combination with neural networks. A case study: Prediction of the HPLC retention index

The molecular structure forms one of the basic ingredients used to model the relationship between chemical compounds and their properties. If neural networks are used as a technique to build this model, the representation of the molecular structure has to fulfil some constraints. This study first draws up an inventory of a number of commonly used representations and discusses their applicability in combination with neural networks. Many representations are not easy to obtain or lack information on the atom types. Therefore an alternative molecular structure representation, i.e. type distance counting (TDC), has been introduced which is based on distances between atom pairs and the respective atom types. The potential of TDC has been demonstrated by virtue of a simple experiment in which neural networks were used to predict the boiling points of a small set of alkanes and alkenes represented by TDC. Next, with a case study, i.e. modelling the relationship between chemical compounds and the high performance liquid chromatography retention index with neural networks, a comparison of a number of molecular structure representations has been made. The TDC outperformed all other representations used. Then, for this particular problem, the neural network performance has been determined for a number of possible TDC representations. Finally, these results have been compared with the outcome of experiments using the same data sets of chemical compounds and retention indices, but other technique/representation combinations: neural networks/fragment coding, and expert systems/fragment coding. Neural networks combined with the TDC representation appeared to give the best overall performance.     

An algorithm for velocity calculation in close proximity to body surfaces in panel methods

To model incompressible flow over a body of arbitrary geometry when using vortex methods, it is necessary to construct an irrotational field to impose the impermeability condition at the surface of the object. In order to achieve this impermeability, this paper uses a boundary integral equation based on the single-layer representation for the velocity potential. Specifically, we formulate this exterior Neumann problem in terms of a source/sink boundary integral equation. The solution to this integral equation is then coupled with an interpolation procedure which smoothes the transition between near-wall and interior regimes. We describe the numerical scheme embedding this strategy and discuss its accuracy and efficiency. For validation purposes, we consider the potential and vortical flow over a circular cylinder, for which an analytical solution and the commonly used method of images are available.     

Extending Superquadrics with Exponent Functions: Modeling and Reconstruction

Superquadrics are a family of parametric shapes which can model a diverse set of objects. They have received significant attention because of their compact representation and robust methods for recovery of 3D models. However, their assumption of intrinsical symmetry fails in modeling numerous real-world examples such as the human body, animals, and other naturally occurring objects. In this paper, we present a novel approach, which is called extended superquadric, to extend superquadric's representation power with exponent functions. An extended superquadric model can be deformed in any direction because it extends the exponents of superquadrics from constants to functions of the latitude and longitude angles in the spherical coordinate system. Thus, extended superquadrics can model more complex shapes than superquadrics. It also maintains many desired properties of superquadrics such as compactness, controllability, and intuitive meaning, which are all advantageous for shape modeling, recognition, and reconstruction. In this paper, besides the use of extended superquadrics for modeling, we also discuss the recovery of extended superquadrics from 3D information (reconstruction). Experiments on both realistic modeling and extended superquadric fitting are presented. Our results are very encouraging and indicate that the use of extended superquadric has potential benefits for the generation of synthetic images for computer graphics and that extended superquadric also is a promising paradigm for shape representation and recovery in computer vision.     

词汇语义表示研究综述

构建能够表达语义特征的词语表示形式是自然语言处理的关键问题。该文首先介绍了基于分布假设和基于预测模型的词汇语义表示方法,并给出目前词表示方法的评价指标;进而介绍了基于词汇表示所蕴含的语义信息而产生的新应用;最后,对词汇语义表示研究的方法和目前面临的问题进行了分析和展望。
     

多语种文本自动生成系统中句子规划部分知识表示模型的构造

提出了一种多语种文本自动生成系统中句子规划阶段的知识表示模型,它以句子结构类、句法规则和语义词典确定文本的具体形式,并详细介绍了该知识表示模型的结构及其匹配准则。     

The law as a dynamic interconnected system of states of affairs: a legal top ontology

In this paper, an abstract model of the law is presented that has three primitives: states of affairs, events and rules. The starting point of the abstract model is that the law is a dynamic system of states of affairs which are connected by means of rules and events. The abstract model can be regarded as a top ontology of the law, that can be applied to legal knowledge representation. After an elaboration of the three primitives, the uses of the abstract model are illustrated by the analysis of central topics of law. Then we discuss heuristic guidelines for legal knowledge representation that are suggested by the abstract model. The paper concludes with a comparison with related work. The appendix contains a formalism for the abstract model.     

Attributed linear intermediate representations for retargetable code generators

This paper illustrates the usefulness of an attributed prefix linear intermediate representation for compiler code generation. In separating the machine-independent and machine-dependent aspects of a compiler, we discuss the advantages and disadvantages of an attributed linear intermediate representation with respect to tree-structured intermediate representations. Some of these issues are relevant to fundamental questions of compiler structure with particular emphasis on retargetability. We discuss our implementation experience using this linear intermediate representation with a table-driven code generation scheme for a variety of target architectures.     

A model for reasoning about persistence and causation

Reasoning about change requires predicting how long a proposition, having become true, will continue to be so. Lacking perfect knowledge, an agent may be constrained to believe that a proposition persists indefinitely simply because there is no way for the agent to infer a contravening proposition with certainty. In this paper, we describe a model of causal reasoning that accounts for knowledge concerning cause-and-effect relationships and knowledge concerning the tendency for propositions to persist or not as a function of time passing. Our model has a natural encoding in the form of a network representation for probabilistic models. We consider the computational properties of our model by reviewing recent advances in computing the consequences of models encoded in this network representation. Finally, we discuss how our probabilistic model addresses certain classical problems in temporal reasoning (e. g., the frame and qualification problems).     

Computing the Numerical Scale of the Linguistic Term Set for the 2-Tuple Fuzzy Linguistic Representation Model

When using linguistic approaches to solve decision problems, we need the techniques for computing with words (CW). Together with the 2-tuple fuzzy linguistic representation models (i.e., the Herrera and MartÍnez model and the Wang and Hao model), some computational techniques for CW are also developed. In this paper, we define the concept of numerical scale and extend the 2-tuple fuzzy linguistic representation models under the numerical scale. We find that the key of computational techniques based on linguistic 2-tuples is to set suitable numerical scale with the purpose of making transformations between linguistic 2-tuples and numerical values. By defining the concept of the transitive calibration matrix and its consistent index, this paper develops an optimization model to compute the numerical scale of the linguistic term set. The desired properties of the optimization model are also presented. Furthermore, we discuss how to construct the transitive calibration matrix for decision problems using linguistic preference relations and analyze the linkage between the consistent index of the transitive calibration matrix and one of the linguistic preference relations. The results in this paper are pretty helpful to complete the fuzzy 2-tuple representation models for CW.      

基于FLAME的通用ABMS模型表示方法改进

模型表示方法是Agent建模过程中需要解决的最重要问题之一。通用ABMS模型表示方法的研究能够降低Agent建模工具的使用门槛,从而使其在跨学科领域的研究中发挥巨大作用。FLAME作为现有平台中的佼佼者,不仅完成度高,而且在模型表示、代码生成及可视化显示等方面具有突出的优势。在这样的前提下,选取FLAME平台与XML作为通用模型表示方法的基础,在继承其简洁、完备、并行化等优点的基础之上,通过挖掘其潜在的技术弱点与改进空间,提出了针对性的解决方案。最后,通过一个Agent模型实例验证了改进方案的可行性。     

基于语言模型的预训练技术研究综述

预训练技术当前在自然语言处理领域占有举足轻重的位置。尤其近两年提出的ELMo、GTP、BERT、XLNet、T5、GTP-3等预训练模型的成功,进一步将预训练技术推向了研究高潮。该文从语言模型、特征抽取器、上下文表征、词表征四个方面对现存的主要预训练技术进行了分析和分类,并分析了当前自然语言处理中的预训练技术面临的主要问题和发展趋势。     

Optimal representation of non-stationary random fields with finite numbers of samples: A linear MMSE framework

In this article we consider the representation of a finite-energy non-stationary random field with a finite number of samples. We pose the problem as an optimal sampling problem where we seek the optimal sampling interval under the mean-square error criterion, for a given number of samples. We investigate the optimum sampling rates and the resulting trade-offs between the number of samples and the representation error. In our numerical experiments, we consider a parametric non-stationary field model, the Gaussian–Schell model, and present sampling schemes for varying noise levels and for sources with varying numbers of degrees of freedom. We discuss the dependence of the optimum sampling interval on the problem parameters. We also study the sensitivity of the error to the chosen sampling interval.     

天然糖甙类化合物^13C—NMR位移范围规律的计算机解析方法

本文从糖甙类化合物的数据特点出发,用较严格的数学推导总结出原则上适用于任何类型的~(13)C-NMR 图谱化学位移范围规律计算机自动解析的方法,并从多方面说明了该方法的研究及实现对实际的~(13)C-NMR 图谱解析工作的意义和重要作用。     

An improved algorithm for gray image representation using non-symmetry and anti-packing model with triangles and rectangles

Although the triangle non-symmetry and anti-packing model (TNAM) representation for gray images is an effective image representation method, there is still much space left for optimization. In this paper, inspired by the optimization idea of the packing problem, we proposed an improved algorithm for gray image representation using the non-symmetry and anti-packing model with triangles and rectangles (NAMTR). By comparing the representation algorithm of the NAMTR with those of the TNAM and the popular linear quadtree, theoretical and experimental results presented in this paper show that the former can greatly reduce the number of sub-patterns or nodes and simultaneously save the data storage much more effectively than the latter, and therefore it is a better method to represent gray images. Representation method of the NAMTR, as envisaged in this paper, shows a very strong promise, and it is valuable for further theoretical research and potential business foreground, such as reducing storage space, increasing transmission speed and improving pattern match efficiency.     

An improved algorithm for gray image representation using non-symmetry and anti-packing model with triangles and rectangles

Although the triangle non-symmetry and anti-packing model (TNAM) representation for gray images is an effective image representation method, there is still much space left for optimization. In this paper, inspired by the optimization idea of the packing problem, we proposed an improved algorithm for gray image representation using the non-symmetry and anti-packing model with triangles and rectangles (NAMTR). By comparing the representation algorithm of the NAMTR with those of the TNAM and the popular linear quadtree, theoretical and experimental results presented in this paper show that the former can greatly reduce the number of sub-patterns or nodes and simultaneously save the data storage much more effectively than the latter, and therefore it is a better method to represent gray images. Representation method of the NAMTR, as envisaged in this paper, shows a very strong promise, and it is valuable for further theoretical research and potential business foreground, such as reducing storage space, increasing transmission speed and improving pattern match efficiency.     

THE CHALLENGE OF DEEP MODELS,INFERENCE STRUCTURES,AND ABSTRACT TASKS

Abstract

An attempt is made to see recent developments in knowledge representation, clustered around Minsky 's theory of frames, and the more classical modeling of natural processes, originating in Newton's mechanics, in a unified perspective. Defining knowledge as homomorphic representation, we view classical models as capable of providing behavioral knowledge but lacking in capability to represent important structural knowledge and meta-knowledge. He discuss a framework, inspired by system theoretic concepts, for combining structural and behavioral representation in a form suitable for goal-directed model construction. We conclude by drawing several implications of this approach for system design, artificial intelligence, modeling, and simulation.     

A hierarchical representation for heterogeneous object modeling

A hierarchical representation for heterogeneous object modeling is presented in this paper. To model a heterogeneous object, Boundary representation is used for geometry representation, and a novel Heterogeneous Feature Tree (HFT) structure is proposed to represent the material distributions. HFT structure hierarchically organizes the material variation dependency relationships and is intuitive in modeling different types of material gradations. Based on the HFT structure, a recursive material evaluation algorithm is proposed to dynamically evaluate the material compositions at a specific location. Such a hierarchical representation guarantees complex material gradations and the user's design intent can be intuitively represented. Example heterogeneous objects modeled with this scheme are provided and potential applications are discussed.     

Analysis neutral data structure for GD&;T

The fundamental issue for automatic geometric tolerance analysis is the representation model, which should, in conjunction with CAD models, accurately and completely represent the GD&T specification according to the GD&T standards. Furthermore, such a representation model should facilitate GD&T validation and tolerance analysis. Most GD&T representation models proposed so far are specific to the tolerance analysis method. Common tolerance analysis methods are min/max chart, Monte Carlo simulation and multivariate regions. This paper will propose a semantic GD&T model, which can be used for any of these methods. The model is a super constraint-tolerance-feature-graph (SCTF-Graph). This paper will demonstrate how the SCTF-Graph model can represent all the tolerance types in the standards, and can contain all the information that is needed for tolerance analysis: nominal geometry (i.e. trimmed features in this research), constraints, tolerances, degrees of freedom (DoFs) to be controlled, assembly hierarchy, and their respective inter-relationships. This paper will discuss the content of the model, how it can be automatically created from the CAD model containing GD&T information (e.g. attributed B-Rep model), and the implementation of such a model, along with some case studies.