网络上SIR型传播的随机建模与极限定理

复杂网络本身及其上的传播过程是最近以来的研究热门,但是多数的工作都只有宏观模型,没有微观基础.本文中,我们将边与边之间的连接与个体的状态变化看作为基本的泊松流,从微观角度构造了网络上SIR型传播的随机模型.对任意有限的固定时间区间,在热力学极限下,从随机系统可以得到一个确定性动力系统—所谓的"水动力极限".这样,本文为这一领域的唯象的动力学模型建立了一个合理的微观基础.同时,我们得到了收敛速度的估计.     

高校收费和招生改革的社会影响的博弈模型分析

针对我国高校近年来不断扩大招生规模,提高收费标准等一系列改革举措,本文用微观经济分析的方法,通过建立相应的博弈模型,具体分析了这一系列改革对学生的影响.主要结论一是高校扩招(或提高学费)增大了学生读书的机会成本,同时降低了读书的效用,但只要学生读书的效用严格大于不读的效用,高校就有动力扩大招生或提高学费);二是扩招和提高学费对学生的影响具有动态相对性,其影响力的大小与现阶段二者的相对水平有关;三是在一定条件下扩大招生规模尤其是提高收费水平会使经济条件不同的学生在是否深造和是否选择名校或热门专业上作出完全不同的抉择,产生了教育公平问题.此外还得到了关于高校改革对其自身收益影响的一些结论.     

非时齐马尔科夫链及其轨道的概率构造

给出了一种由局部刻画构造非时齐马尔科夫链的概率方法.以任意时刻之后第一次跳的时间为条件,证明了过程的马尔科夫性,同时得到了转移概率的递归表达式—这种做法使其概率意义得以明确呈现,并且进一步证明了强马尔科夫性等一系列性质.这为数值模拟,即以Monte Carlo生成非时齐Q-过程的随机轨道提供了严格的理论基础.     

异质底层结构对SIR型传播的两面性影响:阈值和爆发规模(英文)

本文的主要目的是研究底层结构对易感一得病—移除(SIR)型传播的影响.除了周知的阈值决定因素λc=〈k~2〉/〈k〉之外,本文发现另一个看起来矛盾的现象:当传染率λ超过阈值λc后,异质结构对于传播具有消极影响.本文对机制作了定量的解释.更进一步,本文引进了"合作性"的概念,来测量异质性对于爆发规模的总体影响,发现"合作性"负相关于底层结构度分布的变异系数.这个结果说明,底层结构的异质性在疾病的开始阶段和传播阶段有不同的作用模式.     

Duplication： a Mechanism Producing Disassortative Mixing Networks in Biology

Assortative/disassortative mixing is an important topological property of a network. A network is called assortative mixing if the nodes in the network tend to connect to their connectivity peers, or disassortative mixing if nodes with low degrees are more likely to connect with high-degree nodes. We have known that biological networks such as protein-protein interaction networks （PPI）, gene regulatory networks, and metabolic networks tend to be disassortative. On the other hand, in biological evolution, duplication and divergence are two fundamental processes. In order to make the relationship between the property of disassortative mixing and the two basic biological principles clear and to study the cause of the disassortative mixing property in biological networks, we present a random duplication model and an anti-preference duplication model. Our results show that disassortative mixing networks can be obtained by both kinds of models from uncorrelated initial networks. Moreover, with the growth of the network size, the disassortative mixing property becomes more obvious.     

Warp of the Invariant Circle and Onset of Chaos in Josephson Junction Equation

The dynamics of the dc and ac driving Josephson junction equation is studied in terms of the two-dimensional Poincaré map. The smooth invariant circle on the phase cylinder in over-damped case a ） 2 loses smoothness as a decreases and becomes a strange attractor eventually. This triggers two kinds of chaos, one occurs in the regions between two Arnold tongues and the other occurs within the tongues.