Description: 几篇关于神经网络泛化的论文.几篇关于神经网络泛化的论文-several neural network generalization of the papers. Several of the generalization of the neural network papers Platform: |
Size: 632832 |
Author:曾志伟 |
Hits:
Description: 递归神经网络程序，用于递归神经网络的训练，并可以进行泛化。-Recurrent neural network procedure for recurrent neural network training and generalization can be carried out. Platform: |
Size: 1024 |
Author:杨丽 |
Hits:
Description: 小波神经网络程序，用于小波神经网络的训练及泛化求解。-Wavelet neural network procedure for wavelet neural network for solving the training and generalization. Platform: |
Size: 2048 |
Author:杨丽 |
Hits:
Description: rbf神经网络应用于系统辨识,比BP网络具有较好的泛化能力,学习速度快,辨识效果好!-rbf neural network applied to system identification, than the BP network has better generalization ability, learning speed and better recognition! Platform: |
Size: 1024 |
Author:liyan |
Hits:
Description: 人工神经网络(ANN)的泛化特性是神经网络最重要的特性,同时也是最不容易保证的特性。本程序对改进泛化的神经网络算法以及新兴的机器学习算法——支持向量机算法进行研究,-Artificial Neural Network (ANN) the generalization characteristics of neural networks are the most important characteristics, but also not easy to guarantee the most features. This procedure for improving the generalization of neural network algorithm, as well as the emerging machine learning algorithms- Support Vector Machine algorithm research, Platform: |
Size: 7168 |
Author:王旭 |
Hits:
Description: 支持向量机和BP神经网络虽然都可以用来做非线性回归，但它们所基于的理论基础不同，回归的机理也不相同。支持向量机基于结构风险最小化理论，普遍认为其泛化能力要比神经网络的强。为了验证这种观点，本文编写了支持向量机非线性回归的通用Matlab程序和基于神经网络工具箱的BP神经网络仿真模块，仿真结果证实，支持向量机做非线性回归不仅泛化能力强于BP网络，而且能避免神经网络的固有缺陷——训练结果不稳定。
-Support Vector Machine and BP neural network, even though there can be used to make non-linear regression, but they are based on the theoretical basis for the different, the mechanism of regression is not the same. Support vector machine based on structural risk minimization theory, generally considered the generalization ability of neural networks than the strong. To test this view, the paper prepared by non-linear regression support vector machine procedures and based on a common Matlab neural network toolbox of BP neural network Platform: |
Size: 3072 |
Author:孙准 |
Hits:
Description: 摘 要： 介绍BP算法神经网络由线拟舍方法，并借助MATLAB工具箱函数将它运用于方解石色散特性研
究，通过拟合效果图，误差曲线，误差范数反映BP神经网络的优越性，体现BP算法较高的预测能力和良好的泛化能
力，并且可以自动地确定数学模型．精确度高，原理也较简单，尤其对复杂的输入输出系统具有更好的效果。-Abstract： Curve fitting method of BP neural network was introduced and applied in the model of the dispersion of calcite
crystals by MATLAB tools．The results show that BP algorithm has high forecasting capacity and good generalization capacity
in three areas：the map of curve fitting，the deviation curve and the error norm．BP neural network can automatically identify
mathematical model，which has higher precision，and its principle is relatively simple．So it is a very good tool for complex
input-output system． Platform: |
Size: 293888 |
Author:zhenzhen |
Hits:
Description: 支持向量机和BP神经网络虽然都可以用来做非线性回归，但它们所基于的理论基础不同，回归的机理也不相同。支持向量机基于结构风险最小化理论，普遍认为其泛化能力要比神经网络的强。为了验证这种观点，本文编写了支持向量机非线性回归的通用Matlab程序和基于神经网络工具箱的BP神经网络仿真模块，仿真结果证实，支持向量机做非线性回归不仅泛化能力强于BP网络，而且能避免神经网络的固有缺陷——训练结果不稳定。-SVM and BP neural networks, although non-linear regression can be used to do, but they are based on different theoretical basis, the return mechanism is not the same. SVM based on structural risk minimization theory, generally considered the generalization ability of neural networks than strong. To test this view, a support vector machine of this writing the general non-linear regression procedures and based on Matlab neural network toolbox of the BP neural network simulation module, the simulation results confirm that support vector machines do not only the generalization ability of non-linear regression in BP network, and neural networks to avoid the inherent shortcomings- the training results unstable. Platform: |
Size: 11264 |
Author: |
Hits:
Description: 针对神经网络的瓦斯预测模型存在的泛化性能差且存在易陷入局部最优的缺点，提出了
基于最小二乘支持向量机(LS-SVM)时间序列瓦斯预测方法．由于标准最小二乘支持向量机
(L孓SVM)要求样本误差分布服从高斯分布，且标准LS-SVM丧失鲁棒性与稀疏性等特点，提出
了基于加权LS-SVM的瓦斯时间序列预测的方法，从而提高了标准L孓SVM模型的鲁棒性．其
中时间序列的嵌入维数与延迟时间采用了微熵率最小原则进行选取，在此基础上给出了基于加
权L孓SVM实现多步时间序列预测的算法实现步骤．最后利用MATLAB 7．1对其进行仿真研
究，通过鹤壁十矿1个突出工作面的瓦斯涌出数据实例对模型进行了验证．结果表明，加权
SVM模型比标准的L§SVM明显提高了鲁棒性，可较好地实现时间序列数据的多步预测．-The neural network gas prediction model is poor in generalization performance and
easy in fafling into the local optimal value．In order to overcome these shortcomings，we pro—
pose the time series gas prediction method of least squares support vector machine(L§SVM)．
However，in the LS-SVM case，the sparseness and robustness may lose，and the estimation of
the support values iS optimal only in the case of a Gaussian distribution of the error variables．
So，this paper proposes the weighted L孓SVM tO overcome these tWO drawbacks．Meanwhile，
the optimal embedding dimension and delay time of time series are obtained by the smallest dif—
ferential entropy method．On this basis，multi-step time series prediction algorithm steps are
given based on the weighted LS-SVM．Finally，the data of gas outburst in working face of Hebi
lOth mine iS adopted to validate this model．The results show that the predict effect of shortterm
the face gas emission is better using the weighted LS-SVM model than using Platform: |
Size: 490496 |
Author:wanggen |
Hits:
Description: ：根据公交站点客流集散量，选用合适的ＢＰ神经网络构建公交车辆调度形式的神经网络预报
模型．运用ＢＰ神经网络Ｍａｔｌａｂ工具箱设计的基本方法与过程，将ＢＰ网络模型引入公交车辆的调
度方案研究，计算结果表明，ＢＰ模型应用于公交车辆调度形式预测中具有较高的预测精度和良好
的泛化能力-according to the bus station passenger flow distribution quantity, choose suitable BP neural network construct public transport vehicle scheduling form of neural network prediction
Model. Using BP neural network Matlab toolbox basic design method and process, the BP network model of bus into the adjustment
Degree plan research, the calculation results show that the BP model is applied to public transport vehicle scheduling form prediction is of high precision and good
The generalization ability of the Platform: |
Size: 140288 |
Author:段建丽 |
Hits:
Description: Using MATLAB tools for MLP NNs (e.g., newff, …), design a two-layer feed-forward neural network as a classifier to categorize the input geometric shapes.
- The snapshot and bitmap of shapes are given:
- Training shapes: shkt.bmp
- Training patterns: trn.txt (each shape is in a 125*140 matrix)
- Test shapes: shks.bmp
- Test patterns: tsn.txt (each shape is in a 125*140 matrix)
- Since the dimension of inputs is too high (17500-dimensional), it is not possible to apply them directly to the net. So, … .
- Try the number of hidden neurons to be at least.
- Do training of NN until all training patterns are truly classified.
- To examine the generalization ability of your NN after training,
a) Apply it to the test patterns and report the accuracies.
b) Add p noise (p=5, 10, …, 75) to the training shapes (only degrade the
black pixels of the shapes) and report in a plot the accuracy versus p.-Using MATLAB tools for MLP NNs (e.g., newff, …), design a two-layer feed-forward neural network as a classifier to categorize the input geometric shapes.
- The snapshot and bitmap of shapes are given:
- Training shapes: shkt.bmp
- Training patterns: trn.txt (each shape is in a 125*140 matrix)
- Test shapes: shks.bmp
- Test patterns: tsn.txt (each shape is in a 125*140 matrix)
- Since the dimension of inputs is too high (17500-dimensional), it is not possible to apply them directly to the net. So, … .
- Try the number of hidden neurons to be at least.
- Do training of NN until all training patterns are truly classified.
- To examine the generalization ability of your NN after training,
a) Apply it to the test patterns and report the accuracies.
b) Add p noise (p=5, 10, …, 75) to the training shapes (only degrade the
black pixels of the shapes) and report in a plot the accuracy versus p. Platform: |
Size: 3072 |
Author:fatemeh |
Hits:
Search - generalization neural network matlab