小神经

The biggest area was 18cm×13cm,and the smallest was 4.5cm×10cm.there were 6 cases osteomyelitis with bone exposure,4 cases muscle tendonexposure.donor-area direct suture or plant skin.resultsall flaps survived,several cases flap were fat and clumsy and need make them thin in second intention.conclusionthe clinical result of repairing soft tissue defects in lower part of shank, the foot, with antidromic island flap providing blood supply by suralis nerver is reliable.the operation is sinple, without sacrifice of major arteries, low infection to foot, convenient for application.

其中骨髓炎骨外露6例，肌腱外露4例。供区直接缝合或植中厚皮片。结果：术后皮瓣完全成活，个别病例术后较臃肿，待二期将皮瓣修薄。结论：应用腓肠神经营养血管逆行岛状皮瓣修复小腿中下段及足部创面的临床效果可靠，手术操作简便，不牺牲小腿主要血管神经，对足部影响较小，便于推广开展。

The factors which should be considered in BP training are found and program is compiled by C++ Builder. Good agreements are obtained by training 30 runs of lower concentration, middle concentration and hyproconcentration flow. The conjunctional functions are memoried and used to pr...

找出了进行BP神经网络训练时应考虑的水沙要素，采用C ++Builder语言编制的程序，对 3 0组高、中、低含沙量的水槽试验资料进行训练，发现训练值与实测值符合较好，说明得到的连接权能反映实际情况；用该连接权对 4组试验数据进行了预测，预测结果与实测值相差较小，表明用BP神经网络方法确定挟沙能力的可行性。

The implantation nerve regenerated Meissner corpuscle in grafting skin,which were distributed in dermatoplasty borderline.The volume of re-generative Meissner corpuscles was58%of normal,relative gray degree was80%,and the regenerative rate of Meissner corpuscles was60%,40%and35%respectively in skin of palm,arm and eyelid.

植入的神经能再生触觉小体，且小体多分布于植皮区的边缘，体积为正常组58%，相对灰度为正常组80%，主要呈单轴突支配；各类皮肤中触觉小体的再生率不同，上睑、前臂、手掌的再生率分别是35%，40%，60%。

The initialized parameters of neural networks are selected, which is directed by rough set theory, and the networks are trained by using a probability-based BP algorithm, which minimizes system errors.

首先，利用粗糙集理论对神经网络初始化参数的选择和确定进行指导，赋予各参数相关的物理意义；然后，以系统输出误差最小化为目标对粗糙神经网络进行训练，使其满足性能要求。

Using BP algorithm of feed-forward neural network in fuzzy neural network, the small disjunction samples are further processed.

采用模糊神经网络中的前向神经网络BP算法对小分裂样本进行进一步处理。

Some SAR image eigenspaces are constructed for different aspects based on independent component analysis. An independent neural network is trained for each of eigenspace of different aspects. The trained neural networks are used to recognize SAR targets, and their results are combined through D-S evidence reasoning.

通过独立分量分析构造若干个方位角的SAR图像特征空间，为每一特征空间各训练一个模糊极小极大神经网络用于分类，然后利用D-S证据理论集成各个神经网络的输出。

Notice that the solvng ELP problem is equivalent to determine a nonlinear mapping,using the nature that BP neural networks has the parallelism and the capabilities of intrinsic nonlinear mapping,we propose a methods of using back error propagation of ne...

而求解线性规划问题的许多有效方法如单纯形法，Karmarkar's方法等却无法直接用于推广的线性规划问题的求解，注意到推广的线性规划问题等价于一个非线性映射，利用BP神经网络的本质并行性和高度非线性处理能力，该文提出了一类求解推广的线性规划问题的BP神经网络方法，它具有网络构建简单，求解精度高，计算量小等优点，实验结果表明其高效性。

In this paper chip shapes were recognized by using of radbas neural network. We put forward area ratio feature, Euler number feature, etc. geometry feature of chip shape image and thing of those features as inputting vector of neural network. Adopt radbas neural network and training the network using RLS.

研究了径向基函数神经网络在硬质合金刀具切屑形态图像识别中的应用，提出了面积比、欧拉数、分散度等硬质合金刀具切屑形态图像的几何特征，以上述特征作为神经网络的输入矢量，利用径向基函数网络，采用了递推最小二乘法训练该网络。

The neural stem cells can be isolated from human embryonic cerebral tissue . The ratio of cell clone formation from smaller gestational age is higher than that from elder gestational age .The differentiation of neural stem cells from hindbrain is easier than that from forebrain.

从人胚脑组织的前、后脑中分离出神经干细胞；较小胚龄的脑组织原代培养时细胞克隆的形成比例较高；后脑组织分离的南京医科大学硕士学位论文神经干细胞比前脑组织来源的更容易分化。

Firstly, two covering hypersphere algorithms were proposed, referred to as minimum volume covering hypersphere algorithm and moving medium centers covering hypersphere algorithm, to heuristically and rapidly select the promising hidden center vectors from all the training samples. Then the particle swarm optimization algorithm combined with recursive orthogonal least square algorithm was used to further prune the initial hidden centers and optimize the control parameters.

为在实际应用中加速优化速度，降低优化计算复杂度，首先提出了最小体积覆盖超球和移动平均中心覆盖超球两种算法，用于启发式的从训练样本中快速的初选出径向基概率神经网络的隐中心矢量，然后将粒子群优化算法和 ROLSA 结合进一步优选隐中心矢量，同时优化核函数控制参数，实现径向基概率神经网络的全结构优化。

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