在算法上

Previous work on this problem has typically assumed that the channel conditions between the base station and the users are governed by a stationary stochastic process.

但在实际情况中，这样的平稳假设并非总是合理的。研究表明基于平稳假设的调度算法，如流行的M-LWDF和EXP-rule算法，在非平稳信道上的性能下降明显。

Based on this,temporal logic formula is used to express the properties that the data flow must satisfy,before the verification,an algorithm transforming the UML Statecharts model into reachable state transition graph is given,finally the correctness of the data flow is verified by model checking technique.

在此基础上，使用时序逻辑公式表示数据流所需满足的性质，在验证数据流的正确性之前，给出了将它的UML Statechart图模型转化为可达状态迁移图的算法，最后通过模型检测算法验证数据流的正确性。

This article used the car license pictures which photographed .under each kind of actual transportation environment for the above algorithm to carry on the experiment, The results of experiments show that the methods that we designed on the one hand enhanced car license processing accuracy, on the other hand reduced the car license process time in the certain degree, therefore even more suited applies to the actual LPRS.

本文使用在各种实际交通环境下拍摄的车牌图片对上述算法进行了试验，试验结果表明本文设计的车牌处理算法一方面提高的车牌处理的正确性，另一方面在一定程度上减少了车牌处理时间，因此更加适合应用于实际的车牌识别系统中。

In this paper, an unconstraint, continuous, twice-differentiable and strictly convex optimization is obtained by utilizing the Huber function to remove the l_1 -norm in the objective of the primal one, which is called the y- approximation of the original primal optimization.

在此基础上，提出一种能够直接在原始空间中求解支持向量机的近似最优解的快速学习算法，并对算法的收敛性、复杂度以及解的性质进行了分析。4。

In order to improve the success ratio and velocity of iris location, an iris location algorithm based on morphological operators was developed.

为了提高在图像质量较差时的虹膜定位性能，改善虹膜定位的成功率和速度，在对Daugman算法改进的基础上，提出了一种基于形态学算子的虹膜定位算法。

The sub-filters of Interacting Multiple-Model algorithm are based on Kalman filtering, which requires the accurate statistical characteristics of noise. Unfortunately, it is usually unknown and only the approximative models can be obtained, which may result in decrease of IMM performance.

交互多模型算法的子滤波器都是基于Kalman滤波的，它要求知道精确的噪声统计特性，然而在许多情况下噪声信号的统计特性是未知的，只能得到噪声信号的近似模型，这在一定程度上降低了IMM算法的跟踪精度。

The high vertical resolution is achieved at a larger sampling distance when interpolation algorithm is adopted to reconstruct the 3D profile and the conicoid fitting technique is used to calibrate for the system error in the data processing.

在数据处理上，采用了内插值三维重构算法，以及二次曲面拟合的系统误差校正算法，可以在较大的采样间距下获取较高的测量精度。

Rough error processing that frequently used to digital filtering was taken as the research object,on the basis of difference method commonly used in the past and profile method of fractile method,a synthesis algorithm-fractile method based on difference method was proposed.The algorithm for processing the rough error data was perfected .The method had been used in a detection system and result is satisfactory.

对数字滤波系统中经常使用的数据处理算法做了综合处理，对于数据采集过程中所得数据做粗差判断处理为主要研究对象，在过去常用的差分法和分位数分布图法的基础上提出了一种综合处理方法-基于差分法的分位数方法，对系统中的粗差数据处理算法做了更完善的探讨，该方法已在实际检测系统中应用，效果显著。

The algorithm first used KNNModel to build multiple clusters on a given dataset and chose few clusters for each class as representatives to construct a hieratical coding matrix in training phase, and then the matrix was used to train each single classifier.

该算法在训练阶段先通过KNN模型算法在数据集上构建多个同类簇，选取各类中最具代表性的簇形成层次编码矩阵，然后再根据编码矩阵进行单分类器训练。

At the same time, the paper researches the parallelable of dynamic program, and discusses two kinds of parallel algorithm on SMP computer. After that, the paper give a parallel algorithm on cluster of SMP computers, and discuss how to divide the data getting more effective execution of parallel algorithm.

论文也讨论了动态规划算法的可并行性，研究了两种在SMP计算机上实现的动态规划并行方法，分析了在SMP机群系统中，如何通过分割数据的方法实现机群系统上算法的并行优化。

For a big chunk of credit-card losses; the number of filings (and thus charge-off rates) would be rising again, whether

年美国个人破产法的一个改动使得破产登记急速下降，而后引起了信用卡大规模的亏损。

Eph. 4:23 And that you be renewed in the spirit of your mind

弗四23 而在你们心思的灵里得以更新