时间序列

GPS时间序列与强震预测研究

The study of the relation between the GPS time series and the strong shock

Rough集挖掘时间序列的研究

Research of Mining Time Series with Rough Sets

IDS中的黑客信息提取GPS水平位移时间序列地震短期信息特征分析

Analyzing the characteristics of short-term anomalous information of GPS time-series data before earthquake

在本文中将我们改进的圆形反向传播网络模型（ImprovedCircularBackPropagation&ICBP）应用于时间序列预测，进行了单步和多步时间序列预测研究。

In this article , we applied our improved circular back-propagation ( ICBP ) network to single step and multi-steps time series prediction respectively .

具体用到的分析工具有SWOT分析、时间序列分析方法等等。

The analyzing tools include SWOT Analysis , Time Series Analysis , etc.

基于改进的RBF网络的混沌时间序列预测

Improved RBF Network for Chaotic Time Series Prediction

再利用Lyapunov指数的矩形阵算法，计算出交通流时间序列的最大Lyapunov指数。

The maximum Lyapunov Exponent was get from the time series by the matrix algorithm .

由时变ARMA模型描述的一类非平稳时间序列的特性

The Characteristics of Nonstationary Time Series Described by Time-varying ARMA Models

SA信号时间序列分析

SA signal time series analysis

基于结构化类比的时间序列预测算法研究及其在PTA共沸精馏塔建模中的应用

Research on Structured-Analogy-Based Prediction Algorithm for Time Series and Its Application on PTA Solvent System

这种方法以接收信号的功率谱密度函数为时间序列，利用最小二乘格型自适应滤波器经由AR建模而得到高分辨率的时间延迟估计。

This method uses the eross power spectral density ( PSD ) function of two received signals as a time sequence .

SVR在混沌时间序列预测中的应用

The Application of SVR to Prediction of Chaotic Time Series

实践证明，Rough集理论作为一种处理模糊和不确定性问题的有效工具，对于时间序列数据的挖掘同样也是有效的。

Practice proves that rough set theory , as an effective tool to deal with vagueness and uncertainty , is also effective to the time series data mining .

对于这种具有复杂的非线性组合特征的时间序列，直接应用GM（1,1）灰色模型往往精度不高。

For such a suite with the character of complicated non-linear combination , the forecasting results by GM ( 1,1 ) model are not satisfied .

通过对矿井瓦斯涌出量时间序列的模糊分形处理，用BP神经网络对影响因素间的非线性关系进行拟合。

After the time series fuzzy fractal processing of the mine gas emission quantity , the non linear relations of the influence factors were combined with BP neural network .

阐述了基于Rough集的时间序列数据的挖掘策略，重点讨论了时间序列数据中的时序与非时序信息的获取问题。

This paper proposes time series data mining strategy based on a rough set . It mainly discusses the acquisition of time-dependent and time-independent information from time series data .

OLS的一个显著特点是不依赖有关时间序列的先验知识。

OLS is independent of a priori knowledge about the segmented time series .

主要阐述了如何利用时间序列分析法中的复杂度法对PCM混沌编码的复杂度进行分析。

Analysis of complexity of PCM chaotic code by using measure of complexity in time sequence analysis is presented .

第三章是用反卷积分析事件相关fMRI时间序列数据的理论部分。

Chapter Three is the theoretical part of using deconvolution analyzing time series datas of event related fMRI .

空间运动目标雷达散射截面（RCS）序列为非平稳时间序列，常用时间序列分析方法很难对其进行分析和特征提取。

RCS of space object is nonstationary time series , it is difficult to do feature extracting using common time series analysis methods .

文中将最小二乘支持向量机算法应用于混沌时间序列预测中，并同BP网络及RBF网络的预测结果进行了比较分析。

This paper applies Least Squares Support Vector Machine ( LS-SVM ) to chaotic time series prediction , and compares the prediction results with BP network and RBF network .

另一处改进是本论文的关键工作，把手势抽象成为一种时间序列，非监督地对其进行聚类、建模和识别。在熵最小化的基础上，设计了一种基于HMM的分级聚类方法。

Based on entropy minimization , a novel HMM based hierarchical clustering method was invented , aiming at the grouping of temporal sequences .

希尔伯特-黄变换（HHT）是近年来发展起来的一种新的时间序列信号分析方法。

Hilbert-Huang Transform ( HHT ) is a newly developed powerful method for nonlinear and non-stationary time series analysis .

通过对前馈神经网络时间序列数据预测网络模型的建立方法及预测方法讨论，基于BP网络对股票数据进行实际预测。

Establishing of prediction network model in multiplayer feedforward neural network ′ s time series prediction and the prediction design measures are discussed , based on BP network the stock data are forecasted .

计算了模型产生的股价时间序列的Lyapunov指数和关联维，并对其进行主分量分析。

The Lyapunov exponent and correlative dimension of the stock price time series created from the model are calculated , and a principal component analysis is carried out .

由于渔业资源评估中补充量的剧烈变动、亲体量的测量误差以及时间序列的偏差常常使亲体补充量（SR）关系模型的确定存在很大偏差问题。

Variations in environmental variables and measurement errors often result in large and heterogeneous deviations in fitting fish stock-recruitment ( SR ) data to an SR statistical model .

去趋势波动分析（DFA）适宜于研究各类非稳态时间序列的长程幂函数相关性。

Detrended fluctuation analysis ( DFA ) is fit for studies on the long-range exponential correlations of non-stationary time serial .

利用该模型对岸桥电机振动烈度时间序列分别进行1步和4步预测，并与自回归（AR）模型的预测值进行比较。

The one-step and four-step forecastings of the motor vibration severities time serial are made based on the models , and the forecasting results are compared with those using AR model .

从Lorenz和Chen′s系统获得多维时间序列，分别对加不同噪声水平的时间序列检验方法的有效性，并比较不同噪声水平对方法的影响。

We obtain differently multidimensional time series from Lorenz and Chen ′ s systems , and test the above method by the time series with different noise level .

目的：利用SAS程序实现ARIMA模型，探讨ARIMA预测模型在季节性时间序列资料分析中的应用。

Objective : To establishment the SAS procedure of ARIMA Model and to investigate the application of ARIMA predictive model in seasonal time series .