高效选取线性独立的原子特征构建精确机器学习势函数

机器学习势由于具有与第一性原理计算相当的准确性,且低得多的计算成本,在原子模拟中极具前景. 然而原子机器学习势的可靠性、速度和可迁移性在很大程度上取决于原子构型的表示. 适当地选取用作机器学习程序输入的描述符是一个成功的机器学习表示的关键. 本文发展了一种简单有效的方法,可以基于训练数据固有的相关性,从大量待选的描述符中自动选取一组最佳的线性独立原子特征. 通过对几个具有较少冗余线性独立嵌入密度描述符的基准分子构建嵌入原子神经网络势的应用,证明了这种新方法的有效性和准确性. 该算法可以大大简化原子特征的初始选取,并极大地提高原子机器学习势的性能.     

Independent continuous mapping for topological optimization of frame structures

Based on the Independent Continuous Mapping method (ICM), a topological optimization model with continuous topological variables is built by introducing three filter functions for element weight, element allowable stress and element stiffness, which transform the 0-1 type discrete topological variables into continuous topological variables between 0 and 1. Two methods for the filter functions are adopted to avoid the structural singularity and recover falsely deleted elements: the weak material element method and the tiny section element method. Three criteria (no structural singularity, no violated constraints and no change of structural weight) are introduced to judge iteration convergence. These criteria allow finding an appropriate threshold by adjusting a discount factor in the iteration procedure. To improve the efficiency, the original optimization model is transformed into a dual problem according to the dual theory and solved in its dual space. By using MSC/Nastran as the structural solver and MSC/Patran as the developing platform, a topological optimization software of frame structures is accomplished. Numerical examples show that the ICM method is very efficient for the topological optimization of frame structures.The project supported by the National Natural Science Foundation of China (10472003) and Beijing Natural Science Foundation (3042002) The English text was polished by Yunming Chen.     

基于两类切比雪夫多项式的循环矩阵行列式的计算

利用多项式因式分解的逆变换,结合循环矩阵和切比雪夫多项式的特殊结构,首先研究第三类和第四类切比雪夫多项式的通项公式,并给出第三类、第四类切比雪夫多项式的关于行首加r尾r右循环矩阵和行尾加r首r左循环矩阵的行列式的显式表达式,最后给出算法实施步骤.     

浅谈分子对称性中独立的对称元素

通过图示和矩阵两方面来阐述独立的对称元素和非独立的对称元素之间的区别与联系,利用比较法帮助读者正确理解独立对称元素的内涵。     

改革重点：建立现代新闻制度

现代新闻制度是媒体民有、新闻自主、监督官员、保护公民的新闻制度。所谓媒体民有、新闻自主,是指媒体主要由民间出资、民间所有、民间经营、民间分利,传媒进行独立的报道、自由的评论、充分的竞争、不受权力或者资本的控制。实行媒体自由创立、备案登记制度,禁止利用公共财政资金办报办电视台,推动公共媒体的独立运作,限制大资本的持股比例,允许外资进入传播事业,     

基于独立成分分析的高光谱图像异常检测

针对高光谱图像中背景及目标先验知识未知条件下的异常目标检测问题,提出了一种基于独立成分分析(ICA)的异常探测算法.首先估计原始数据的虚拟维(VD)以确定要分离的独立成分个数,在此基础上进行快速独立成分分析(FastICA),然后基于平均局部奇异度选择含异常信息较多的独立成分,最后使用丰度量化算法得到异常目标的丰度图像...     

考虑交通出行惯例的双向行人流模型研究

推广了Baek等人最近提出的一个双向行人流模型,提出了两种改进策略,并从行人平均速度-密度关系、行人空间分布密度和位置分布等方面进行了数值分析. 研究发现,引入的两个新策略不仅可以提高行人流的平均速度,而且可以提高道路系统(尤其是道路中央区域)利用率,减轻拥堵状况,有效避免严重堵塞的发生. 改进的策略对行人的心理特点和行为特性等方面考虑更加全面,而且可以较好地模拟高密度的双向行人流. 关键词： 元胞自动机模型 双向行人流 交通惯例     

Uniform estimate for the tail probabilities of randomly weighted sums

Several authors have studied the uniform estimate for the tail probabilities of randomly weighted sumsa.ud their maxima. In this paper, we generalize their work to the situation thatis a sequence of upper tail asymptotically independent random variables with common distribution from the is a sequence of nonnegative random variables, independent of and satisfying some regular conditions. Moreover. no additional assumption is required on the dependence structureof {θi,i≥ 1）.     

Schur Convexity for Two Classes of Symmetric Functions and Their Applications

For x =(x1, x2, ···, xn) ∈ Rn+∪ Rn-, the symmetric functions Fn(x, r) and Gn(x, r) are defined by r1 + xFij n(x, r) = Fn(x1, x2, ···, xn; r) =x1≤iij1i2···ir ≤n j=1and r1- xGij n(x, r) = Gn(x1, x2, ···, xn; r) =,x1≤i1i2···ir ≤n j=1ij respectively, where r = 1, 2, ···, n, and i1, i2, ···, in are positive integers. In this paper,the Schur convexity of Fn(x, r) and Gn(x, r) are discussed. As applications, by a bijective transformation of independent variable for a Schur convex function, the authors obtain Schur convexity for some other symmetric functions, which subsumes the main results in recent literature; and by use of the theory of majorization establish some inequalities. In particular, the authors derive from the results of this paper the Weierstrass inequalities and the Ky Fan's inequality, and give a generalization of Safta's conjecture in the n-dimensional space and others.     

Quasi-maximum likelihood estimators in generalized linear models with autoregressive processes

The paper studies a generalized linear model(GLM)y_t = h(x_t~T β) + ε_t,t = l,2,...,n,where ε_1 = η_1,ε_1 =ρε_t +η_t,t = 2,3,...;n,h is a continuous differentiable function,η_t's are independent and identically distributed random errors with zero mean and finite variance σ~2.Firstly,the quasi-maximum likelihood(QML) estimators of β,p and σ~2 are given.Secondly,under mild conditions,the asymptotic properties(including the existence,weak consistency and asymptotic distribution) of the QML estimators are investigated.Lastly,the validity of method is illuminated by a simulation example.