算法

Algorithms in specific, the proposed blocktree structure, the algorithm to reduce memory requirements and broaden the scope of application of the algorithm.

在具体的算法中，提出了块树的结构，减少了算法所需的内存，拓宽了算法的应用范围。

So the priority that combines packet delay and dynamic queue length is obtained. A packet scheduling algorithm based on this priority scheme is proposed for the input memory switch. The scheduling algorithm supports both packet delay and packet queue length. The application of this scheduling algorithm in Crossbar and Clos networks is presented.

在此优先级的基础上，提出了输入缓存交换结构中能够兼顾分组时延与分组队长的分组调度算法，将此调度算法用于单Crossbar交换单元和Clos网络，并提出了在Clos网络中按照优先级均衡负载的路由算法。

The encryption algorithms ensure the security of message transfer, which among the clients and the authentication servers.

其中，最常用的就是对称加密算法和公钥加密算法，加密算法可以保证用户和认证服务器间的数据传输保密和安全。

In order to solve the problem of chart projection transformation among all projections in shiphandling simulator,using the theory of power series spreading out and the method of trigonometric power series calculating back, a direct and reverse solving algorithm models of meridian arc length , and Gauss projection and Mercator projection transforming algorithm are introduced in this paper.

为了解决船舶操纵模拟器中不同投影方式海图的投影变换问题，应用幂级数展开理论和三角级数回求方法，给出了子午线弧长、等量纬度的直接正反解算法模型及高斯投影与墨卡托投影的相互转化算法。这些算法具有非常高的计算精度，避免了迭代运算，在船舶操纵模拟器中得到广泛应

Our algorithm contains two steps: firstly, a small number of dependence tests are initiated to restrict the feasible search space and reduce calculation complexity; then, in order to ensure accuracy, a MDL metric search in "Delta Form" boosts time performance and efficiency of BN learning.

首先，通过小运算量的独立性测试获取系统结构信息，限制可行解搜索空间，降低了BN学习算法的计算复杂度；然后，采用基于MDL理论的BN学习算法对出现"德尔塔"的局部进行搜索和MDL评分，在充分保证结果精度的前提下，提高了算法的时间性能。

The uncertainty of travel time of road network is depicted by means of scenario union firstly,then a path finding modal applied minimax theorem is constructed for obtaining most reliable path of all scenarios.

该方法通过场景集描述网络旅行时间的不确定性，应用Minimax理论构建求解所有场景下可靠路径的数学模型，并设计了问题求解算法，分析了算法的时间复杂性，最后通过典型算例对算法进行了验证。

For solving the scalable problem and getting optimization solution, we propose the GA-MWPL-DC-ST algorithm which aims at resolving minimum weighted path latency degree-constrained spanning tree based on the two layers ALM model.

本文在具有超节点的双层应用层组播模型基础上，提出了利用遗传算法构建出度受限最小带权路径延时生成树的生成算法GA-MWPL-DC-ST，利用该算法可在超节点上对双层组播树进行分布式构建，从而将求最优解问题的巨大计算量分担到多个超节点上。

In the algorithm, radial-basis function neural network was utilized as forward model, and genetic algorithm was used to solve the optimization problem in the inverse problem, which can avoid the local minimum solution possibly encountered in the iterative inverse algorithm based on gradient descent and consequently achieve the global optimized solution of the inverse problem.

在该算法中，径向基函数神经网络用作前向模型，遗传算法用于求解逆问题中的优化问题，其优点是能够避免基于梯度下降法的迭代逆算法中可能遇到的局部最小问题，并能得到逆问题的全局最优解。

The way of Arnoldi is based upon the principle of Galerkin, which is closed to the solution of the linear numerical system. Here, we call the solution as Generalized Minimum Residual.

基于Galerkin原理的Arnoldi算法是求解这种线性代数方程组的近似算法，以下称这种方法为广义极小残余算法。

Second, taking risk taste coefficient into the model solve the problem of without risk taste in model. Third, it builds up a new model based on the thought of GA and mistiness mathematics as the fit function in GA.

第二，以遗传算法中的目标函数的方式建立了基金投资组合模型，以风险调整的收益的最大值作为遗传算法中的目标函数，应用遗传算法进行了求解。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。