Asymptotic Information-Theoretic Detection of Dynamical Organization in Complex Systems

The identification of emergent structures in complex dynamical systems is a formidable challenge. We propose a computationally efficient methodology to address such a challenge, based on modeling the state of the system as a set of random variables. Specifically, we present a sieving algorithm to navigate the huge space of all subsets of variables and compare them in terms of a simple index that can be computed without resorting to simulations. We obtain such a simple index by studying the asymptotic distribution of an information-theoretic measure of coordination among variables, when there is no coordination at all, which allows us to fairly compare subsets of variables having different cardinalities. We show that increasing the number of observations allows the identification of larger and larger subsets. As an example of relevant application, we make use of a paradigmatic case regarding the identification of groups in autocatalytic sets of reactions, a chemical situation related to the origin of life problem.     

Symmetry specified enumeration of substituted derivatives: an easy solution to the complex problem

Several sophisticated methods to solution of symmetry specified enumeration problems are available in the modern literature. In this paper we propose a simple technique that allows one to manually compute the exact numbers of fixed-symmetry derivatives for a given structure either with inclusion or ignoring the substitution patterns. The basic idea of the method suggested consists in the derivation of Pólya-like cycle indices for the automorphism groups of specially constructed orbit partition graphs; the expansion of these indices and subsequent simple calculations result in the desired numbers of substituted derivatives with achiral substituents. Limitations of the new technique (and a method suggested earlier) depend on the relevance of the orbit partitions for particular subgroups of the point symmetry group. For illustration purposes, the results obtained for the prismane (D _3h ) and adamantane (T _d ) structures are discussed. In the former case the numbers of substituted derivatives can be found for all subgroups of the D _3h group, whereas in the latter case these numbers can be determined for eight out of eleven subgroups of the T _d point symmetry group. This work is based on the text of the lecture presented by the authors at the 5th All-Russia Conference on Molecular Modeling (Moscow, April 2007). The paper deals with the methodology and detailed treatment of applied aspects related to solution of enumeration problems for substituted derivatives with prescribed symmetry groups. Unlike the known methods of symmetry specified enumeration, the technique suggested is simple enough and may be regarded as generalization of the Pólya methodology, which is widely used by chemists. Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 2, pp. 227–245, February, 2008.     

基于哈希理论和线性近邻传递反馈的乳腺X线图像肿块检索方法

在乳腺X线图像肿块检测中存在较高的假阳性率,通过基于内容的肿块检索,将待判定肿块与已确诊肿块进行相似性分析,可有效降低假阳性率.本文提出了一种结合可区分锚点图哈希和线性近邻传递的乳腺图像肿块检索方法.针对传统锚点图哈希在相似度定义中没有考虑病理相关性的问题,引入病理类别至锚点图哈希图像相似度计算,提出了可区分锚点图哈希以重新表示图像.利用线性近邻传递作为相关反馈技术,基于图像底层特征表达与图像高层语义间的学习机制,实现交互式肿块图像检索.采用北京大学人民医院乳腺中心提供的临床图像作为实验数据,实验结果表明,引入病理类别的可区分锚点图哈希图像表达在肿块相似性分析上优于传统锚点图哈希.相比于现有方法,本文提出的方法在肿块检索性能上得到明显提高.     

基于自适应核学习相关向量机的乳腺X线图像微钙化点簇处理方法研究

采用自适应核学习相关向量机方法, 结合形态学滤波和Kallergi分簇标准, 研究了乳腺X线图像中微钙化点簇的处理. 首先将微钙化点检测看作一个监督学习问题, 然后应用自适应核学习相关向量机作为分类器判断图像中每一个位置是否为微钙化点并采用形态学处理滤除干扰噪声, 最后对获得的微钙化点采用Kallergi标准进行分簇. 为提高运算速度, 在微钙化点检测时将整个图像分解为多个子图像并行运算, 实现了一种基于自适应核学习相关向量机的微钙化点簇快速处理方法. 实验结果和分析表明, 自适应核学习相关向量机方法算法性能优于相关向量机方法, 特别是实现的快速方法能进一步降低微钙化点簇的处理时间. 关键词： 乳腺X线图像 微钙化点簇 相关向量机 自适应核学习     

基于局域相关向量机回归模型的小尺度网络流量的非线性预测

基于非线性时间序列局域预测法与相关向量机回归模型, 本文提出了局域相关向量机预测方法, 并应用于预测实际的小尺度网路流量序列. 应用基于信息准则的局域预测法邻近点的选取方法来选取局域相关向量机回归模型的邻近点个数. 对比分析了局域相关向量机预测法、前馈神经网络模型与局域线性预测法对网络流量序列的预测性能, 其中前馈神经网络模型的参数采用粒子群优化算法来优化. 实验结果表明: 邻近点优化后的局域相关向量机回归模型能够有效地预测小尺度网络流量序列, 归一化均方误差很小; 局域相关向量机回归模型生成的时间序列具有与原网络流量时间序列相一致的概率分布; 局域相关向量机回归模型的预测精度好于前馈神经网络模型的与局域线性预测法的. 关键词： 小尺度网络流量 非线性时间序列预测方法 局域预测法 相关向量机回归模型     

基于FHW的火箭发动机环境载荷评估

针对固体火箭发动机环境载荷的诸多不确定性,提出了环境载荷FHW量化评估算法.对比分析了火箭发动机的典型环境条件及其影响,采用熵值法确定了指标权重,基于FHW方法建立了火箭发动机的环境载荷评估模型,根据环境因素的分布特征给出评估指标的隶属函数,并获取其相应的灰色关联系数.算例验证了不同环境载荷因素对火箭发动机的影响程度,结果表明FHW评估策略和算法简单易行、精度高,在装备环境载荷量化评估中具有参考和应用价值.     

基于0-1稀疏循环矩阵的测量矩阵分离研究

当前压缩感知中测量矩阵的优化是测量阶段和重构阶段采用同一矩阵的事前优化。采用了以行变换为主的测量矩阵优化算法和过渡矩阵将压缩感知的测量矩阵和重构矩阵相分离,在测量阶段采用单像素相机的0-1稀疏矩阵,在重构阶段采用近似矩阵,这是区别于传统思路的测量数据和测量矩阵的事后优化方法。理论分析和实验结果表明,优化矩阵的性能好于稀疏循环矩阵,近似矩阵和优化矩阵具有相近的性能。研究成果降低了测量矩阵工程设计和实现的难度。     

基于改进型相关向量机的高光谱图像分类

相关向量机(RVM)高光谱图像分类算法是一种基于贝叶斯概率模型的监督机器学习算法,其分类精度较高、测试时间较短。然而算法本身存在训练时间随着训练样本增加直线上升、分类效率整体降低等问题。针对这种情况,提出一种基于改进型相关向量机(VRVM)的高光谱图像分类算法。本算法在传统概率模型中引入一个新的分布,使得计算复杂度较高的积分运算可近似地拆分成两个较为简单的对数和形式。实验结果表明,VRVM高光谱图像分类算法的总体分类精度和相关向量的数量与RVM基本相同,但训练时间随样本数的增加有明显的减少。     

“高分子材料合成创新实验”线上教学探索与实践

Based on the platform of "QQ share screen + video live broadcast", through the combination of "online theory teaching", "development of home alternative experiment", "game type" experimental operation and "virtual simulation experiment", we carried out the online teaching of polymer material synthesis innovation laboratories, integrating "theory + practice" and "virtual + reality", which is also a good opportunity for labor education. These explorations and practices are expected to provide some references for online teaching of chemical laboratory courses during the COVID-19 epidemic.