二样本问题

Sample Size Determination is commonly encountered in modern medical studies and statistical analysis for two independent binomial experiments.

样本容量的确定在现代生物医学研究以及在对两个独立的二项实验进行统计分析时，是经常遇到的一个问题。

The traditional Canonical Correlation Analysis based image recognition methods always encounter the Small Sample Size problem,which is due to the size of sample and less than the dimension of sample.In order to solve this problem,a new supervised learning method called Two-Dimensional CCA（2DCCA） is developed.

针对传统典型相关分析（Canonical Correlation Analysis，CCA）的图像识别中出现的小样本（Small Sample Size，SSS）问题，提出二维典型相关分析（Two-Dimensional CCA，2DCCA）。

Taking the Guyang county of Baotou city as a test area, the ETM ＋ multi spectral data and CBS pan data are merged by the method of Principal Components Analysis.

为解决此问题，笔者以包头固阳县为研究区，利用主成分变换的方法，对多源遥感影像（ETM＋多光谱数据和中巴资源二号卫星全色波段数据）进行融合处理；同时，在分类中，采用监督分类辅以目视解译分类法相结合的混合分类法，改进训练样本选取方法，进行变化信息的提取，同时给出了易混淆地物的遥感解译标志。

Firstly, the sectional data from 27 companies in communications and transportation industry, which are listed before 1999, are linearly regressed. We find their capital structure and firm performance are irrelated significantly. Then the section data are quadricly regressed, but the result is also disillusionary, which indicates that the capital structure of listed companies has not the best range only with the sectional data from 1999 to 2003. Finally, in allusion to the problems in the securities market and illogicality of the capital structure, we put forward some methods to optimize the capital structure for listed companies from four facts.

文章首先对交通运输仓储行业1999年以前上市的27家公司的截面数据进行线性回归，发现在分析时段1999年至2003年，样本公司的净资产收益率和单位经济增加值与资产负债率之间线性不相关；然后文章又对该行业上市公司的截面数据进行二次函数回归，同样也没有得到令人满意的结果，这说明仅就1999年至2003年五年的截面数据而言，上市公司资本结构不存在最优区间；最后，文章对我国证券市场存在的问题以及该行业上市公司资本结构的不合理性提出了改进的设想，并从四个方面提出了优化我国上市公司资本结构的建议。

Research on the related attribute of evidence, influence and method of dealing with is carried out. Result of experiment demonstrates that the method can lead to the reliable decision effectively. Aiming at conflicting problem of decision making, fusion strategy of BPA reassign is adopted in coordinator of GDSS. Aiming at contradictory of decision-making, the process of decision-making is divided into two stages that are learning and decision-making. Each Agent is trained by proper information given by training example to eliminate contradictories among decision-making Agents and correct mistakes that ensure Agents cooperate effectively in the stage of cooperating decision.(3) In order to overcome the flaw of current GDSS that can only support static problem solving, adopting the method that combing Multi-Decision-Making-Agents Reinforcement Learning and the metempirical knowledge learning, the paper puts forward Reinforcement Learning and optimization under dynamic environment.

提出采用证据理论的证据合成作为协调器中决策Agent的融合策略，对群体决策支持系统的协调器进行决策融合过程中可能出现的情况进行了详细的分析，研究了证据的相关性及其在决策融合过程中的影响，提出了相关证据的处理方法，实验结果表明，该方法可有效保证决策结果的可靠性；针对决策过程中的冲突问题，提出协调器采用基本信任指派二次分配的协调策略；针对决策过程中的矛盾，提出将决策过程分为学习和决策两个阶段，各决策者Agent根据有限的训练样本所给出的正确信息进行训练，消除决策者Agent之间的矛盾和修正决策者Agent自身的错误，从而保证合作决策阶段实现有效的合作决策。

Do you know, i need you to come back

你知道吗，我需要你回来

Yang yinshu、Wang xiangsheng、Li decang,The first discovery of haemaphysalis conicinna.

1〕 杨银书，王祥生，李德昌。安徽省首次发现嗜群血蜱。