搜索资源列表
一个主要用于预测和分类的源程序
- 开发环境:Matlab 简要说明:程序包含的设置包括:列数、样本总个数、建模样本数、预报因子数、预报对象数、学习因子、 动量因子、训练次数、总体误差、隐节点数。-development environment : Matlab brief descr iption : program includes the installation including : number, the total sample number, the number of samples modeling
trnn
- 神经网络训练,应用matlab7NN包,用一个隐藏层使用5折交叉验证。-Training the Neural Network This scr ipt is something that I did for a course at Uni. It uses the Neural Networking package provided with MatLab 7 unfortunately I m not sure if it s available with the earlier ve
bp3
- 三层前馈神经网络的BP算法。程序具有以下功能: (1) 允许选择各层节点数; (2) 允许选用不同的学习率η; (3) 能对权值进行初始化,初始化用[-1、1]区间的随机数; (4)允许选用单极性和双极性两种不同Sigmoid型转移函数。 -Three-tier feed-forward neural network BP algorithm. Procedures have the following functions: (1) allows to choose the
SGA40
- 用标准遗传算法解决电力系统最优化问题,粒子使用的是40个节点的大规模计算问题-Using standard genetic algorithm to solve power system optimization problems, particle using 40 nodes of a large-scale computing problems
nn_nodes_optm
- 计算最优化网络节点传输路径的neural network 模型,测试通过4个nodes-Calculation of the optimal transmission path network node neural network model, the test four nodes
BPNN
- 用BP神经网络实现模糊控制规则为T=int[(e +ec)/2]的模糊神经网络控制器。可以改变隐层节点数和学习速率。网络训练算法是变学习速率法。-BP neural network with fuzzy control rules for the T = int [(e+ Ec)/2] of the fuzzy neural network controller. Can change the hidden layer nodes and learning rate. Network tra
ga_bp
- 基于gaot工具箱优化BP以实现数据拟合。在nninit.m中修改数据存储的路径以及隐含层节点数的设置(S1),遗传算法与bp的参数修改在ga_bp.m中进行。-Toolbox gaot Optimize based on BP in order to achieve data fitting. At nninit.m modify the path of data storage as well as the hidden layer nodes of the set (S1), geneti
resourceAllocate
- 神经网络的一种资源构造算法,用matlab实现,在训练过程中不断增加网络的节点数直到满足性能要求。-Neural network construction algorithm of a resource, using matlab implementation, in the training process of increasing the network nodes until you meet the performance requirements.
shouxieshibie
- 上午上课在稿纸上画了半天,整理下思路,下午动手写的,所有模块已大概成型,本着照顾后学者,以及资源共享的原则,贴上核心的bp算法部分,已经封装好了,使用可以直接调用。三层,基本三层就可以解决我们的这些简单的问题,输入为8×8,输出4,隐层数目conts num=8 还可进一步封装,const输入层节点,输出层节点。-Morning class in the writing paper drew a long time, under the idea of finishing the afterno
TimeSeriesPredictionUsingSupportVectorRegressionNe
- 为了选择神经网络的最好结构以及增强模型的推广能力,提出一种自适应支持向量回归神经网络(SVR—NN)。SVR—NN 用支持向量回归(SVR)方法获得网络的初始结构和权值, 白适应地生 成网络隐层结点,然后用基于退火过程的鲁棒学习算法更新网络结点疹教和权 主。 SVR—NN有很 好的收敛性和鲁棒性,能抑制由于数据异常和参数选择不当所导致的“过拟合,’现象。将SVR—NN 应用到时间序列预测上。结果表明,SVR.NN预测模型能精确地预测混沌时间序列,具有很好的 理论和应用价值。-Ab
Cellular-Neural-Network
- 细胞神经网络(CNN)是一种和人类神经网络非常相似的并行计算模型,各个邻接节点间有不同的通信。在本程序中A模型是反馈矩阵,B是控制矩阵。-Cellular neural network (CNN) is very similar to the human neural network model of parallel computation, all adjacent nodes have different communication. A model of this process is
8-nodes
- 8节点的前推回代法潮流程序,可以计算配电网的潮流。-8 nodes matlab power flow
ANN
- 编写BP神经网络系统,可适应多节点输入输出的人工神经网络训练,即可以自由定义输入节点的个数,隐层节点个数,输出节点个数以及样本的个数。-Preparation of BP neural network system can be adapted to multi-node input and output of the artificial neural network training, that is free to define the number of input nodes, hid
IC
- 该文档为BP算法在智能控制中的应用例子及其说明。BP网络输入、输出节点数可选;隐含层数及隐含层节点数可选,最大训练次数及退出训练误差可选,数据文件可选,动量因子可选,学习速率可选,激活函数类型可选,网络实现功能(分类、拟合)可选。开发环境为MATLAB。-This document is BP algorithm in intelligent control application examples and their descr iptions. BP network input and ou
Simscape-Bryson_Denham
- Simcape是用MATLAB编写的开源的软件包,它使用直接配置法,供多区间最优控制问题使用。该软件包采用具有LGL配置点的Legendre伪谱法和具有CGL配置点的Chebyshev伪谱法。-Simcape is an open source software package written in MATLAB that uses direct collocation methods for multiple-interval optimal control problems. This pa
power-flow-analysis
- 基于matlab平台编写程序,分析IEEE118与300节点标准数据,并给出最优潮流。-Matlab platform based on the preparation of the program, with 300 nodes analysis IEEE118 standard data, and gives optimal power flow.
boltz
- matlab平台玻尔兹曼机机器学习,3个节点,线性降温-Matlab platform prepared by the Boltzmann machine learning, three nodes, linear cooling,
rbf
- 自己编写RBF神经网络程序,RBF神经网络隐层采用标准Gaussian径向基函数,输出层采用线性激活函数,其中数据中心、扩展常数和输出权值均用梯度法求解,它们的学习率均为0.001。其中隐节点数选为10,初始输出权值取[-0.1,0.1]内的随机值,初始数据中心取[-1,1]内的随机值,初始扩展常数取[0.1,0.3]内的随机值,输入采用[0 1]的随机阶跃输入(Write your own RBF neural network, RBF neural network hidden layer