交叉验证

针对支持向量机的参数选择问题，文中采用基于k折交叉验证误差的网格搜索法。

The grid search method based on k fold cross validation error for selecting the model parameters of SVM is also presented .

在两个真实数据集上用10层交叉验证对朴素贝叶斯信用评估模型进行了测试，并与五种DavidWest的神经网络个人信用评估模型进行了对比。

They are tested using 10-fold cross validation with two real world data sets , and compared with five neural network models of David West 's.

在两个干扰性不同的样本集上，分别对FAN模型和人工神经网络模型进行交叉验证。

Cross-validations of FAN model and ANNs ( Artificial Neural Networks ) were conducted on two different sample datasets .

基于交叉验证的改进RBF分类器设计

Design of RBF Network Based on cross-validation method

并采用基于统计学习的交叉验证法确定了最佳SVM分类参数。

Then Cross-Validation method which is based on statistical learning is introduced for confirming the best classification parameter . 4 .

同时，运用交叉验证法优化SVM的惩罚参数和核参数，有效提高了预测精度。

And using cross validation to optimize the SVM penalty parameter and kernel parameter , effectively improve the prediction accuracy .

本文建立了基于支持向量机（SVM）的变压器状态评估模型。采用交叉验证法确定SVM分类器参数。

The transformer condition estimate model is constructed based on SVM and the parameter for SVM-based classifier is determined by adopting cross validation method .

V-fold交叉验证和BP神经网络在信用评价中的应用

An Application of V-fold Cross-validation Technique and BP Neural Networks to Credit Analysis

在对结构模型和估计方案进行交叉验证后，对g（AuAg）的空间分布进行估计。

After cross & validating the parameters of the structure model and the estimating project , the spatial distribution of g ( Au / Ag ) was modeled .

分析中还发现Kriging插值模型的选择及Crossin&ValNation交叉验证的检验与系数修正均影响分析工作的结果精度。

Choosing proper Kriging model , verifying of the Cross-Validation and correcting the parameters are important factors influencing accurate analysis .

所构建的CoMFA模型具有较好的交叉验证系数和一定的预测能力。

The CoMFA model had good cross-validated coefficient ( q2 ) and predictive potency .

为了进一步提高诊断准确率，文中运用Libsvm安装包中的交叉验证功能选取SVM的最佳参数c和g，并利用最佳参数来优化模型。

In addition , in order to further improve the diagnostic accuracy , the cross-validation function in the installation package is used to select the best paraments c and g , which can optimize the model .

本文介绍交叉验证参数R~2（q~2）引导的构象选择CoMFA方法，选择化合物的最佳构象。

A cross-validated R2 ( q2 ) guided conformation selection approach of CoMFA studies was proposed .

交叉验证和预测结果表明，所建立的QSPR模型具有良好的稳定性和预测能力。

Cross validation and prediction results show that this QSPR model is of good stability and powerful prediction ability .

通过9次交叉验证实验，结果表明SVM★对车辆数据样本的测试准确率达到了85.59%，其分类性能优于其它分类器。

The testing accuracy to this vehicle dataset reaches 85.59 % by means of 9-fold cross-validation which demonstrates that the classification performance of SVM ~ (★) is superior to those of other classifiers .

在UCI数据集上用分层交叉验证对该算法进行了对比测试，结果表明该算法具有良好的分类精度。

Experiment results show that this algorithm holds a good accuracy of classification by stratification-cross-validation on the sets of UCI . 3 .

针对目标分子柔性大的特点，在比较分子场分析（CoMFA）方法中采用交叉验证相关系数平方R~2引导的构象选择法.对12个皂甙分子的生物活性进行了三维定量构效关系研究。

In this article , the R2 guided region selection CoMFA method was used to investigate 12 saponins for their 3D - QSAR .

通过仿真分析，确定了以径向基核函数（RBF）和交叉验证法选取最佳惩罚因子的ε-SVR算法作为建模的核心算法。

Through the simulation experiment , we selected the radial basis function ( RBF ) as the kernel function , and by the cross validation method we determined the best punishment factor for the ε - SVR algorithm .

最后用BP神经网络建模，采用Leave-n-out交叉验证法对模型进行验证，并讨论了隐含层神经元个数、学习速率、动量因子和学习次数对所建BP网络的影响。

The influences of the number of hidden neurons , learning rate , Momentum , and epochs were discussed in this paper , and the models were validated using leave-n-out cross-validation approach .

选择了高斯核函数和多项式核函数，应用交叉验证的方法对SVM进行训练识别，得到了最优的参数模型，取得了有一定意义的实验结果，为进一步的研究奠定了较好的基础。

Training and identifying SVM by selecting Gaussian kernel function and polynomial kernel function and using cross-validation method , the optimal parameter model is obtained and the significant experimental results are achieved , which lay a good foundation for further studies .

5折交叉验证ROC曲线下面积在0.93以上，平均为0.967，其中，主成分约简结果模型差于其它两种约简结果模型，差异有显著性意义。

The average AUC in 5-fold cross-validation was more than 0.93 , average 0.967 , and that of principle components reduction model was lower than the other two reduction models with statistic significance .

然后分别采用主成分回归（PCR）和偏最小二乘回归（PLS）方法，通过留一法交叉验证选择潜变量个数，建立具有较好预测能力的定量构效关系（QSAR）模型。

A leave-one-out cross-validation method was used to select the number of latent variables for the building of the QSAR models by principal component regression ( PCR ) and partial least square regression ( PLS ) method respectively .

基于软阈值的广义交叉验证准则（GCV）已运用于图像去噪中，但此方法对图像峰值信噪比（PSNR）的改善有限，不能有效保持细节。

The denoising method based on GCV has been used in image denoising , but it is hard to improve PSNR greatly and can ′ t maintain the details of the image well .

在应用SVM分类器之前，先用样本数据进行训练，运用交叉验证和网格搜索技术优化SVM的RBF核函数参数，然后拍摄番茄图像用于SVM分类器识别实验，分析实验结果，并提出了多类识别方向。

Before SVM classifier application , a set of data was used for training , the use of cross-validation and grid search techniques to optimize the SVM RBF kernel parameters . Afterward tomato picture was took for the SVM classifier to recognition experiments .

交叉验证表明，所建模型平均拟合相对误差为0.0063%，平均预报相对误差为0.1210%，该模型可用于提纯塔出口CO2浓度的预测。

The cross experiments indicated that the average matching relative error and the average prediction relative error of this model , was 0.0063 % and 0.1210 % respectively . The model was applied to predict the concentration of CO_2 in the exit of purifying column .

hv-block交叉验证法在率定水文模型参数中的比较研究

Hv - Block Cross-Validation Method for Hydrological Model Calibration

该方法将SVR参数集当作粒子群，以最小化5-fold交叉验证误差作为适应目标，利用PSO强劲的全局搜索能力实现了参数优选。

This method takes SVR parameters as the particle swarm and the minimization of the 5-fold cross-validation error as the adaptation goal . Then it uses PSO with strong global search ability to achieve the parameter optimization .

对原始光谱进行平滑和二阶导数处理，用RBF神经网络法建模，采用Leave-one-out交叉验证法对模型进行验证，并对参数的影响进行了讨论。

The raw spectra were pretreated by the second derivative and smoothing , prediction models were established by using RBF , the models were validated using Leave-one-out cross-validation approach , and the influence of parameters was discussed .

单尾ANOVA分析表中只有农民人均纯收入指标的显著性不佳，而判别分析中交叉验证样本的分类正确率为88.9%，说明分类的结果比较理想。

In the one-tailed ANOVA analysis table only the index " rural per capita net income " had the highest F value and the classification accuracy of Cross-validation samples is 88.9 % in discriminant analysis which show the classification results is ideal .

本文指出了应用分离样本验证法率定水文模型参数的不足，引入并探讨了hv-block交叉验证法的适用性，并与分离样本验证法进行比较。

This paper pointed out the disadvantages of the spilt-sample validation method and introduced the hv-block cross-validation method for the hydrological model calibration .