相依稳健t过程函数型ANOVA模型

随着数据的多元化和复杂化,函数型ANOVA模型已经被越来越多的学者研究并应用到各行各业.为了减弱模型对于离群值的敏感度,本文对于相依的多元响应变量提出了一种相依稳健t过程函数型ANOVA模型.为了刻画变量之间的相依性和保证协方差函数的正定性,本文引入了随机效应变量和卷积方法构造了一个协方差函数.另外,模型中引入了随机效应函数去描述研究对象的个体特征.统计性质,例如稳健性和信息相合性在本文中也得到了相应的研究,并通过数值模拟和实例分析来验证此模型的可行性.     

一类具有扩散和Beddington-DeAngelis反应函数的病毒模型的全局稳定性

研究了一类具有扩散和Beddington-DeAngelis反应函数的病毒模型.通过构造Lyapunov函数,证明了模型的感染平衡点是全局渐近稳定的.     

随机删失单函数型指标模型条件众数估计的渐近正态性

基于单函数型指标模型,构造了该模型下条件密度和条件众数的估计量,研究了 α-混合函数型数据在响应变量随机删失的情况下的条件密度和条件众数估计量的渐近正态分布,用模拟研究说明单函数型指标模型条件众数估计的有效性.     

部分函数线性模型的模型平均方法

部分函数线性模型是一种被广泛研究和应用的模型,其响应变量与一般的随机变量有关,也与函数型的随机变量有关.文章首先利用传统的谱分解方法来表示协方差函数,将部分函数线性模型的函数部分线性化,其次基于Hansen (2007)的Mallows模型平均方法,提出了该模型下的最优权重的选择准则,并证明了模型平均估计量的渐近最优性,此外还考虑了候选模型为两个特殊模型的情况下的模型平均估计量的渐近最优性.最后,进行了模拟研究,并对肉类和玉米样本的近红外反射光谱数据集进行分析,均表明所提出的模型平均方法是有效的.     

具有营养循环和时滞的浮游生物系统的稳定性及Hopf分支

建立并研究了具有营养循环和时滞的浮游动植物模型,模型中描述浮游动植物间的相互作用函数是Holling-Ⅲ型功能反应函数.首先讨论了模型解的正性及有界性,然后分析了系统在无时滞和有时滞两种情况下边界平衡点和正平衡点的局部稳定性,并通过建立适当的Lyapunov函数,讨论了平衡点的全局稳定性.研究表明,随着时滞的增加,系统会出现Hopf分支.     

具有Holling Ⅱ型功能性作用函数的随机恒化器模型的渐近性态

考虑了一类营养的转化率受到随机噪声干扰,具有Holling Ⅱ型功能性反应函数的随机恒化器模型.通过构造Liapunov函数,利用停时、伊藤公式证明了模型正解的全局存在唯一性.研究了模型解的长期渐近性态,主要揭示在不同条件下模型的解围绕其相应确定性模型的各类平衡点的振荡行为.     

双阀值LSTAR模型非线性检验的理论研究及其应用

本文分析—个转换函数为logistic形式的双阀值平滑转换自回归(LSTAR)模型,并对该模型的非线性进行了诊断检验.通过泰勒展开等手段,用一个简单的线性模型逼近非常复杂的非线性的双阀值LSTAR模型,对应的两个模型的零假设被证明完全等价,从而大大简化模型线性性的诊断检验工作.理论研究表明,不同于单阀值STAR模型需要对转换函数进行三阶泰勒展开,双阀值模型只需要对转换函数进行一阶泰勒展开即可,这大大节省了自由度.模拟研究和实际数据分析结果验证了理论结果的有用性.     

二参数广义Rayleigh分布

为了研究某些工程过程的寿命,需要一个能够适应这些过程性质的寿命模型.与简单模型相比,基于寿命分布的广义模型在建模过程中更具有直观的适用性和吸引力.本文介绍一个二参数广义Rayleigh分布,推导出了其矩生成函数和k阶矩的表达式,通过对样本点进行归一化,得到了它的近似分布函数、概率密度函数、生存函数、风险率函数等.在此基...     

典型非线性模型的模糊辨识逼近精度分析

人们根据非线性系统的复杂特性归结了几种具有代表性的非线性模型.而模糊辨识方法是辨识非线性系统的有力工具,本文采用T-S模糊模型对三种常见的非线性模型:Hammerstein模型,Wiener模型和双线性模型进行逼近,并根据仿真数据研究不同的非线性结构对模糊模型逼近精度的影响.仿真实例是在训练和检验数据组数、模型阶数相同的情况下,采用三角形隶属函数,聚类型隶属函数和高斯型隶属函数分别对这三种非线性模型进行逼近能力的研究.     

空间半参数变系数部分线性回归中的分位数估计

本文研究了空间数据变系数部分线性回归中的分位数估计. 模型中的参数估计量通过未知系数函数的分段多项式逼近得到, 而未知系数函数的估计量通过将参数估计量代入模型中并通过局部线性逼近得到. 文中推导了未知参数向量估计量的渐近分布, 并建立了未知系数函数估计量在内点及边界点的渐近分布. 通过Monte Carlo 模拟研究了估计量的有限样本性质.     

Stability of antiperiodic recurrent neural networks with multiproportional delays

In general, a proportional function is obviously not antiperiodic, yet a very interesting fact in this paper shows that it is possible there is an antiperiodic solution for some proportional delayed dynamical systems. We deal with the issue of antiperiodic solutions for RNNs (recurrent neural networks) incorporating multiproportional delays. Employing Lyapunov method, inequality techniques and concise mathematical analysis proof, sufficient criteria on the existence of antiperiodic solutions including its uniqueness and exponential stability are built up. The obtained results provide us some lights for designing a stable RNNs and complement some earlier publications. In addition, simulations show that the theoretical antiperiodic dynamics are in excellent agreement with the numerically observed behavior.     

New results on input-to-state convergence for recurrent neural networks with variable inputs

In this paper, we study a class of recurrent neural networks (RNNs) arising from optimization problems. By constructing appropriate Lyapunov functions, we prove two new results on input-to-state convergence of RNNs with variable inputs. Numerical simulations are also given to demonstrate the convergence of the solutions.     

带有时滞的Clifford值神经网络的全局指数稳定性

研究了带有离散时滞和分布时滞的Clifford值递归神经网络的全局指数稳定性问题.首先运用M矩阵的性质和不等式技巧证明了Clifford值递归神经网络平衡点的存在性和唯一性;然后通过数学分析方法,得到了Clifford值递归神经网络全局指数稳定的判定条件;最后数值仿真例子验证了获得结果的有效性.     

一类递归小波神经网络的稳定性研究

在小波神经网络（WNNs）和递归神经网络（RNNs）的基础上,提出了一类递归小波神经网络（RWNNs）模型,它具有两种网络模型的优点A·D2根据Liapunov渐近稳定理论,对该模型的渐近稳定性进行了研究,并给出了相关的定理和公式．仿真结果表明该模型对非线性动态系统有良好的辨识效果．     

Global exponential stability in Lagrange sense for recurrent neural networks with time delays

In this paper, we study the global exponential stability in Lagrange sense for continuous recurrent neural networks (RNNs) with multiple time delays. Three different types of activation functions are considered, which include both bounded and unbounded activation functions. By constructing appropriate Lyapunov-like functions, we provide easily verifiable criteria for the boundedness and global exponential attractivity of RNNs. These results can be applied to analyze monostable as well as multistable neural networks.     

ON THE PRODUCT OF GTEAUX DIFFERENTIABILITY LOCALLY CONVEX SPACES

A locally convex space is said to be a Gateaux differentiability space (GDS) provided every continuous convex function defined on a nonempty convex open subset D of the space is densely Gateaux differentiable in .D.This paper shows that the product of a GDS and a family of separable Prechet spaces is a GDS,and that the product of a GDS and an arbitrary locally convex space endowed with the weak topology is a GDS.     

ON THE PRODUCT OF GATEAUX DIFFERENTIABILITY LOCALLY CONVEX SPACES

A locally convex space is said to be a Gateaux differentiability space (GDS)provided every continuous convex function defined on a nonempty convex open subset D of the space is densely Gateaux differentiable in D.This paper shows that the product of a GDS and a family of separable Frechet spaces is a GDS,and that the product of a GDS and an arbitrary locally convex space endowed with the weak topology is a GDS.     

ON THE PRODUCT OF GATEAUX DIFFERENTIABILITY LOCALLY CONVEX SPACES

A locally convex space is said to be a Gâteaux differentiability space (GDS) provided every continuous convex function defined on a nonempty convex open subset D of the space is densely Gâteaux differentiable in D. This paper shows that the product of a GDS and a family of separable Fréchet spaces is a GDS, and that the product of a GDS and an arbitrary locally convex space endowed with the weak topology is a GDS.     

Convergence dynamics of stochastic reaction–diffusion recurrent neural networks with continuously distributed delays

Convergence dynamics of reaction–diffusion recurrent neural networks (RNNs) with continuously distributed delays and stochastic influence are considered. Some sufficient conditions to guarantee the almost sure exponential stability, mean value exponential stability and mean square exponential stability of an equilibrium solution are obtained, respectively. Lyapunov functional method, M-matrix properties, some inequality technique and nonnegative semimartingale convergence theorem are used in our approach. These criteria ensuring the different exponential stability show that diffusion and delays are harmless, but random fluctuations are important, in the stochastic continuously distributed delayed reaction–diffusion RNNs with the structure satisfying the criteria. Two examples are also given to demonstrate our results.