您要查找的是不是：
- In this method, rule intensity is defined according to the number of misclassified training samples. 该算法根据误分类训练样本的数量定义规则强度。
- SVM is used to classify, which weakly depends on the quantity and quanlity of training samples. 在分类器设计方面,选用了对样本数量和质量依赖性小的支持向量机。
- Besides that, we adopted a bootstrapping method during network's training, successfully solving the deficiency of non-face training samples. 而改进后的BP网络缩短了学习时间,提高了学习效率,并在一定程度上避免了学习中的局部极小问题。
- The discrimination model is established from the training samples using BP algorithm,and then the samples is distinguished from the well-trained. 利用BP算法对训练样本进行学习,确定判别模型,根据已训练好的神经网络对样本进行判别。
- In the training of the neural network model (NNM) of the plant and the neural network controller (NNC), training samples are got from the state function of the plant. 在训练实现对象模型的网络和实现控制器的网络时,由状态方程产生训练样本。
- In this paper the training samples, training method of neural network and the way combined with ADRC is analyzed, and the valuable conclusion is obtained. 文中对神经网络的训练样本、方法及其与自抗扰控制器结合的方式进行了分析和讨论，并得出了有益的结论；
- Experiment result shows that as reserving typical samples and reducing training samples, the generalization performance and training efficient of the classifier are guaranteed. 仿真结果证实，由于保留了典型样本，减少了训练样本数量，从而保证了分类器的性能且训练效率较高。
- The algorithm can obtain the better classifiers by using less training samples,so it leads to more generalization and less training samples than the other learning models. 该算法能用较少的训练样本获得更佳的分类器,因此它的推广能力较好,且对训练要求的样本数也大大下降。
- In order to prevent causing network incorrectly incline with one of fault type after training, the number of training samples for every fault should be allocated averagely. 为了防止人为的造成训练后的网络过多的倾向于某一故障类型，各故障类型的训练样本数量不应相差太多。
- Its weighted training error and scaling factor cm is computed (step 3b).The weights are increased for training samples, which have been misclassified (step 3c). 计算错误率和换算系数cm(step 3b).;被错分的样本的权重会增加。
- The training samples of RBF Neural Network are determined by Uniform Experiment Design to accelerate training process and enhance approximate capacity of the neural network. RBF神经网络训练样本的选取则由均匀试验设计确定，以提高样本的代表性并大幅减少样本数量，从而加快网络的训练过程，加强网络的逼近能力。
- Typical fault characteristics are selected as training samples and GA is used to optimize the structure and original weight distribution of BP networks. 该法采用故障的典型特征作为样本训练网络，遗传算法来优化BP网络的结构与初始权值分布的策略。
- Under the nonhomogeneous environment, the performance of the conventional statistical STAP degrades greatly due to lack of sufficient homogeneous training samples. 在非均匀环境下,由于缺乏足够的与待检测样本中干扰独立同分布的训练样本,常规统计STAP方法性能急剧下降。
- The calculation result indicates that the calculation model of water pump performance based on SVM is pithily and the model can be set up with a small amount of train samples. 结果表明，基于支持向量机建立的水泵性能计算模型具有一定的简洁性，只需要知道少量的训练样本就能建立数学计算模型；
- GRNNFA is a hybrid model to perform regression by the General Regression Neural Network (GRNN) based on the kernels obtained from compressing the training samples by Fuzzy ART (FA). GRNNFA是由回归神经网络模型与模糊自适应共振网络模型组合而成的。 先利用模糊自适应共振网络模型从训练样本中压缩出核函数,再通过回归神经网络模型进行预测。
- Simulation tests were carried out with evaluation data given by experts from power supply enterprises in Baoding taken as training samples for BP neural network,and the results are good. 以保定市各县供电企业专家评价数据作为BP神经网络的训练样本,进行仿真试验,得到了满意的结果。
- By computing each density of class in training set and the average density of the whole training set, some samples in the high-density class can be deleted using the training samples reduction method. 通过计算训练样本集中各类别的类别密度及整个训练集的平均密度，去掉高密度类别中的部分样本； 渐进式分类模式模拟人工分类文本的智能化形式，分为按标题分类、按关键段落分类和按全文分类三个层次，尽量减少分析全文的比例。
- Faced with the fact that training samples belonging to normal operation status are much more than ones belonging to abnormal operation status,the weighted support vector machine is presented. 针对污水处理过程运行状态监控中的正常运行状态样本数多而异常运行状态样本数少的特点,提出加权支持向量机方法。
- Firstly，it’s difficult to give out a reasonable tradeoff between the detection rate and the false positive rate.Secondly，the face posture which can be detected depends on the training samples greatly. 通过研究，认为这一算法还存在两个有待改进的地方：一是在检测率和误检率之间难以权衡，二是可检测人脸姿态受训练样本制约。
- The outputs of training samples of the new control model is nearly the same as the ones of checkout samples, and the strong adaptability and accuracy of the two subnets are verified. 新控制模型训练样本的输出结果和检验样本的输出结果相差不大，验证了两个子网具有很强的适应能力和逼近能力。
