Location:
Search - 神经网络，时间序列预测
Search list
Description: 本文针对网络流量的混沌特性，介绍网络流量的相空间重构方法和参数确认的方
法，并通过简单的试验验证理论的可用性。
同时将小波变换和非线性动力学方法相结合研究网络流量的混沌特性，并改进相空
间重构方法。将混沌吸引子投影于小波滤波器向量所张的空间中，并充分利用了小波变
换的去噪优点，将小波变换与相空间重构结合，构建出一个新的重构模型，并用试验证
明其优越性。将小波神经网络混沌时间序列预测方法引入到网络流量预测中，给网络数
据流的预测方法都提供了行之有效的新方法。
Platform: |
Size: 1373594 |
Author: hp_yan@yahoo.cn |
Hits:
Description: 1、该工具箱包括了混沌时间序列分析与预测的常用方法,有:
(1)产生混沌时间序列(chaotic time series)
Logistic映射 - \ChaosAttractors\Main_Logistic.m
Henon映射 - \ChaosAttractors\Main_Henon.m
Lorenz吸引子 - \ChaosAttractors\Main_Lorenz.m
Duffing吸引子 - \ChaosAttractors\Main_Duffing.m
Duffing2吸引子 - \ChaosAttractors\Main_Duffing2.m
Rossler吸引子 - \ChaosAttractors\Main_Rossler.m
Chens吸引子 - \ChaosAttractors\Main_Chens.m
Ikeda吸引子 - \ChaosAttractors\Main_Ikeda.m
MackeyGLass序列 - \ChaosAttractors\Main_MackeyGLass.m
Quadratic序列 - \ChaosAttractors\Main_Quadratic.m
(2)求时延(delay time)
自相关法 - \DelayTime_Others\Main_AutoCorrelation.m
平均位移法 - \DelayTime_Others\Main_AverageDisplacement.m
(去偏)复自相关法 - \DelayTime_Others\Main_ComplexAutoCorrelation.m
互信息法 - \DelayTime_MutualInformation\Main_Mutual_Information.m
(3)求嵌入维(embedding dimension)
假近邻法 - \EmbeddingDimension_FNN\Main_FNN.m
Cao方法 - \EmbeddingDimension_Cao\Main_EmbeddingDimension_Cao.m
(4)同时求时延与嵌入窗(delay time & embedding window)
CC方法 - \C-C Method\Main_CC_Luzhenbo.m
(5)求关联维(correlation dimension)
GP算法 - \CorrelationDimension_GP\Main_CorrelationDimension_GP.m
(6)求K熵(Kolmogorov Entropy)
GP算法 - \KolmogorovEntropy_GP\Main_KolmogorovEntropy_GP.m
STB算法 - \KolmogorovEntropy_STB\Main_KolmogorovEntropy_STB.m
(7)求最大Lyapunov指数(largest Lyapunov exponent)
小数据量法 - \LargestLyapunov_Rosenstein\Main_LargestLyapunov_Rosenstein1.m
\LargestLyapunov_Rosenstein\Main_LargestLyapunov_Rosenstein2.m
\LargestLyapunov_Rosenstein\Main_LargestLyapunov_Rosenstein3.m
\LargestLyapunov_Rosenstein\Main_LargestLyapunov_Rosenstein4.m
(8)求Lyapunov指数谱(Lyapunov exponent spectrum)
BBA算法 -
\LyapunovSpectrum_BBA\Main_LyapunovSpectrum_BBA1.m
\LyapunovSpectrum_BBA\Main_LyapunovSpectrum_BBA2.m
(9)求二进制图形的盒子维(box dimension)和广义维(genealized dimension)
覆盖法 -
\BoxDimension_2D\Main_BoxDimension_2D.m
\GeneralizedDimension_2D\Main_GeneralizedDimension_2D.m
(10)求时间序列的盒子维(box dimension)和广义维(genealized dimension)
覆盖法 -
\BoxDimension_TS\Main_BoxDimension_TS.m
\GeneralizedDimension_TS\Main_GeneralizedDimension_TS.m
(11)混沌时间序列预测(chaotic time series prediction)
RBF神经网络一步预测 - \Prediction_RBF\Main_RBF.m
RBF神经网络多步预测 - \Prediction_RBF\Main_RBF_MultiStepPred.m
Volterra级数一步预测 - \Prediction_Volterra\Main_Volterra.m
Volterra级数多步预测 - \Prediction_Volterra\Main_Volterra_MultiStepPred.m
(12)产生替代数据(Surrogate Data)
随机相位法 - \SurrogateData\Main_SurrogateData.m
2、在matlab环境中首先运行install.m,将工具箱所在路径添加至matlab
3、各子目录下以Main_开头的文件即是主程序文件,直接按快捷键F5运行即可
4、工具箱中所有程序均在Matlab6.5和Matlab7.1环境中调试通过，不能保证在Matlab其它版本正确运行。
5、工具箱中部分功能为试用版，敬请谅解！
6、
作者：陆振波，海军工程大学
欢迎同行来信交流与合作，更多文章与程序下载请访问我的个人主页
Platform: |
Size: 579972 |
Author: niuchao0511 |
Hits:
Description: 采用RBF神经网络的结构、特性和训练算法，根据CPI(消费者物价指数)与其影响因素之间存在的映射关系，应用神经 网络建立了多因素非线性时间序列预测模型。最后通过仿真实验和研究，把RBF神经网络与传统的BP网络预测结果进行比较，结果证明，该模型的预测精确度更高，结果令人满意。
Platform: |
Size: 295664 |
Author: gigixufy@hotmail.com |
Hits:
Description: 基于人工神经网络的时间序列预测matlab源代码
Platform: |
Size: 7167 |
Author: xjxiaojin |
Hits:
Description: 用BP神经网络程序模拟销售预测，能对销售数据进行时间序列预测，采用VC实现-BP neural network simulation sales forecasts, sales data can be right for time series prediction, using VC
Platform: |
Size: 221184 |
Author: 杜昭翼 |
Hits:
Description: 混沌时间序列局域法多步预报模型.doc(有程序下载)
针对混沌时间序列预测中用加权一阶局域法单步预报模型进行多步预报时计算量大且存在误差累积效应的不足，本文提出了基于相空间重构技术的局域法多步预报模型，包括加权一阶局域法多步预报模型和RBF神经网络多步预报模型。对几种典型混沌序列的预测仿真表明，两种模型对混沌时间序列的多步预报均较有效。
-chaotic time series Local Law multi-step prediction model. Doc (with the download) against chaotic time series prediction using a weighted-Local law single-step prediction model multi-step forecast at large calculation error and the cumulative effect of the shortage, In this paper, based on the phase-space reconstruction of local law multi-step prediction model Weighted including a local law-order multi-step prediction model and RBFNN multi-step prediction model. Several typical of the chaotic sequence forecast simulation shows that the two models of chaotic time series multi-step prediction than effective.
Platform: |
Size: 143360 |
Author: 呆雁 |
Hits:
Description: 混沌时间序列预测工具箱，包括了李雅普诺夫指数、分形纬、嵌入纬以及神经网络预测-chaotic time series forecasting tool kit, including the Lyapunov exponent, fractal-wai, Wei and embedded neural network prediction
Platform: |
Size: 355328 |
Author: 四度 |
Hits:
Description: matlab，时间序列，神经网络，预测，控制-matlab, time series, neural networks, prediction, control
Platform: |
Size: 333824 |
Author: 飞鸿 |
Hits:
Description: 灰色预测模型称为CM模型，G为grey的第一个字母，M为model的第一个字母。GM（1，1）表示一阶的，一个变量的微分方程型预测模型。GM（1，1）是一阶单序列的线性动态模型，主要用于时间序列预测。 一、GM（1，1）建模 设有数列 共有 个观察值 对 作累加生成，得到新的数列 灰色理论与模型及在车辆拥有量预测中的应用 灰色神经网络交通事故预测比较
灰色系统(第三版)-projections Gray (No. Third Edition)
Platform: |
Size: 114688 |
Author: fyh |
Hits:
Description: Elman递归神经网络对时间序列的预测代码，做的效果还行，仅供参考-Elman recurrent neural network for time series prediction code, do the results were OK for reference purposes only
Platform: |
Size: 30720 |
Author: chenshengli |
Hits:
Description: 基于长记性特征的时间序列预测模型，很好用，准确度优于普通神经网络，我自己一直在用-Characteristics based on long memory time series forecasting model, the good, the accuracy is better than an ordinary neural network, I have been using
Platform: |
Size: 5120 |
Author: 李文兵 |
Hits:
Description: BP神经网络预测算法MATLAB源程序，用于混沌时间序列预测。(BP neural network prediction algorithm MATLAB source code for chaotic time series prediction.)
Platform: |
Size: 1024 |
Author: 强仔撒
|
Hits:
Description: NAR神经网络 采用matlab编程，用来预测时间序列，(NAR neural network is used to predict the time series)
Platform: |
Size: 1024 |
Author: 雷之夏至
|
Hits:
Description: ELMAN神经网络模型预测精度高于BP神经网络，可以用于非线性时间序列的预测。(The prediction accuracy of ELMAN neural network model is higher than that of BP neural network, and it can be used to predict nonlinear time series.)
Platform: |
Size: 2048 |
Author: yuqian0407 |
Hits:
Description: 小波神经网络的时间序列预测——短时交通流量预测，含源程序和数据(Time Series Prediction Based on Wavelet Neural Network - Short-term Traffic Flow Forecasting)
Platform: |
Size: 4096 |
Author: 潇潇飒飒 |
Hits:
Description: 本文采用小波神经网络进行交通流量预测，短时交通流量存在随机性和非线性因素，影响预测的准确性。传统预测模型难以反映交通流量变化特点，同时传统神经网络易陷入局部极小值，泛化能力差，交通流量预测精度低。为了提高短时交通流量预测精度，提出一种小波神经网络的短时交通流量预测模型。小波神经网络可以对短时交通流量随机性、不确定性进行局部分析，并进行非线性预测，验证了模型的有效性，进行了对比试验。验证结果表明，小波神经网络提高了短时交通流量预精度，预测结果更具应用价值。(In this paper, wavelet neural network is used to forecast traffic flow. There are random and nonlinear factors in short-term traffic flow, affecting the accuracy of prediction. The traditional prediction model is difficult to reflect the characteristics of traffic flow changes. At the same time, the traditional neural network is easy to fall into the local minimum, the generalization ability is poor, and the prediction accuracy of traffic flow is low. In order to improve the accuracy of short-term traffic flow forecasting, a short-term traffic flow prediction model based on wavelet neural network is proposed. The wavelet neural network can analyze the randomness and uncertainty of short-term traffic flow and perform nonlinear prediction. The validity of the model is verified and a comparative experiment is conducted. The verification results show that the wavelet neural network improves the short-term traffic flow pre-precision, and the prediction result has more application value.)
Platform: |
Size: 4096 |
Author: 阳光男孩_LHKF |
Hits:
Description: 基于动态神经网络的时间序列预测模型，可以进行时间序列预测(A time series prediction model based on dynamic neural network can predict time series.)
Platform: |
Size: 1024 |
Author: whateverojbk |
Hits:
Description: 一个关于小波神经网络的时间序列预测模型，完整可运行的代码(A Time Series Prediction Model of Wavelet Neural Network, Complete Running Code)
Platform: |
Size: 3072 |
Author: isxwpan |
Hits:
Description: 小波神经网络代码预测模型，用于时间序列的预测。(This is a source code about wnn. The code is coding by matlab 2016a and it can apply to predict someting based on time series.)
Platform: |
Size: 1024 |
Author: 宏远伟大 |
Hits:
Description: 小波神经网络的时间序列预测——短时交通流量预测，很好
Platform: |
Size: 4047 |
Author: k2008m |
Hits: